Q: Which sportsman electric hammer bot has more weapon Joules: 'Mangi' or 'Helios Sport'? Which one do you like better?

A: Mark J. here: I don't have enough information about 'Mangi' to make even a rough calculation on weapon energy. I've written to Al Kindle for additional info. As a pure guess I'd say 'Helios Sport' hits a little harder, but 'Mangi' is quicker. They did have one head-to-head fight won by 'Helios Sport', but they are both beautifully made top-flight robots.

UPDATE: Al got right back to me with details on Mangi's weapon. According to my calculations, 'Mangi' with 78 Joules is harder hitting than 'Helios Sport' at 50 Joules.

A: Mark J. here: the spreadsheet I have is just a few basic kinematic formulas strung together with a couple 'best guess' fudge factors. I put it together to get some order of magnitude comparisons between electric and pneumatic hammers. It has no documentation, is useable only over a small range of values without recalibration, and it relies on some questionable physics shortcuts. I don't think it would be of any use at all as a design aid.

A full-blown hammer simulator would be kind of a big deal. The motor torque continually changes with increasing speed, there is no good way to calculate the inertia of the armature from available data, and the effect of gravity on the hammer changes non-linearly all along the arc.

My best advice, as has often been repeated on the Ask Aaron webpage, is to use the construction examples provided by other builders as a starting point. In your case I'd carefully examine Helios Sport from Team Cosmos. Best luck!

A: Yes, 'The Butcher' from Inertia Labs had a massive pneumatic powered spinning weapon which was not very successful. Pound for pound, electric batteries store much more energy than compressed gas cylinders. Electric motors are also very efficient (80%) compared to compressed gas engines (15%). Pneumatics are great for quick bursts of power, but not efficient for continuous motion.

Q: Would the hammer get more damage if I increased the arm to 18" and lightened the hammer head to 3oz?

A: Mark J. here: assuming a 180 degree hammer arc and taking a guess at torque losses to armature acceleration:

• the Speed 400 motor geared 25:1 will provide about 5 G's acceleration to a 4 ounce weight on a 12" arm;

• the hammer head will complete a 180 degree arc in about 0.2 second;

• the velocity of the hammer head at the end will be about 31 feet per second;

• the kinetic energy of the hammer head at impact will be around 5 Joules.

Weapon energy is a product of the acceleration provided by the motor and the length of time that acceleration is applied. Changes to the gear ratio, length of the arm, or mass of the hammer will change the weapon acceleration, but will be offset by the change in time available before impact. As long as these design elements remain anywhere near reasonable you'll still get about the same weapon energy output. To get more energy output, you'll need more motor power input.

Q: Is 5 joules strong for a beetle hammer?

A: A strong hammer runs close to 10 Joules per pound of weight class -- about 30 joules for a beetle. Electric hammers just can't put out enough power.

Q: How many joules did 'Helios Sport's hammer have?

A: Mark J. here: 'Helios Sport' is a 30 pound 'sportsman class' robot with a 3 pound hammer on the end of an 18" arm powered by a DeWalt 14.4 volt motor at 18 volts. That works out to about 50 Joules -- comparable to a 5 Joule hammer on a beetle. You can see from the video of 'Helios Sport' vs. 'Bounty Hunter' that the hammer isn't effective.

A: Mark J. here: not quite. The 'weight' input is the constant force pressing down on the end of your lifter. The motor torque needed to offset that force is shown as the green line in the output chart. Note that the torque needed varies with the position of the lifter, but the maximum torque available from your gearmotor is constant. This means that the maximum lift force will vary with lifter position.

Also, the motor torque shown in the graph is the amount needed to just offset the weight on the lifter. Permanent magnet electric motors produce maximum torque at stall (zero speed), so a gearmotor with the torque shown in the graph could hold position against that much weight at that position but could not lift it further. Aim for a gearmotor with about twice the maximum torque shown for the weight you want to lift.

A: 'The Matador' from Inertia Labs claims the most powerful flipper title with 20,000 pounds of lifting force. Team Run Amok's The Gap is a heavyweight that last weighed in at 208 pounds with an empty CO² tank.

Q: Is Matador's flipper more powerful than Toro's?

A: What did I just say? Matador's is the more powerful. Toro's superheavyweight flipper had 7000 pounds of lift. The Matador's 'next generation' flipper had 20,000 pounds of lift in a heavyweight.

Q: Aaron, do you know how many pounds of force Ziggy's lifting arm has? Do you know what psi Ziggy's pneumatic system runs on? Could you give me a link to Ziggy's website if it has one?

A: CM Robotics claims 14,000 pounds of lifter force for 'Ziggy', but the 4-bar flipper mechanism makes comparison with direct-acting flippers like 'Toro' difficult.

Ziggy's flipper runs on high-pressure air, someplace in the 3000 to 5000 psi range.

Some robot teams, such as Ziggy's CM Robotics, don't have websites. It isn't that they aren't capable, it's just that they have a problem with sharing. : ^ )

A: Mark J. here: that number comes from CM Robotics -- I don't know how they are calculating. In a conventional single-pivot lifter it's fairly easy to apply a little geometry to figure the lifter force based on the force available at the pneumatic cylinder (surface area of piston times available pressure), the angle at which the force is applied, and the leverage between the force application point and the point of lift. Ziggy has a four-bar flipper mechanism, which makes the calculation more difficult. They may simply be giving the raw force available at the cylinder.

Actual flipper performance depends not only on the maximum force available, but also on the gas flow capacity of the pneumatic system. Valves and hoses must be large enough to flood the cylinder with an instantaneous burst of high-pressure gas or the 'flip' becomes a gentle 'lift'.

A: The Matador, like all of Inertia Labs' robots, was beautifully made and well thought out. The amount of force in it's weapon was certainly overkill. A little less weight devoted to the weapon and a little more to a more controllable chassis might have worked out better.

It's record was 4 wins and 2 losses -- the last loss to the fearsome 'Warhead' at BattleBots 5.0. I don't remember the match, but there may not have been enough of 'The Matador' left to put back together.

A: There are certainly some gearmotors that will work for your purpose, but it is also a fairly simple task to hack a servo for continuous rotation.

The power you'll need from a gearmotor (or hacked servo) will depend on the length of the lifter arm attached to it. For maximum speed and reliability, select a gearmotor that would just stall with 1.67 times the weightclass on the end of the lifter. Here's the formula to calculate the desired torque:

Example - a gearmotor for a 4" lifter arm on a beetleweight should have:

The B231 231:1 gearmotor running at 12 volts would be a good choice. You could also use the BaneBots 64:1 24mm gearmotor running at 7.2 volts, but it's higher speed could be difficult to control on a lifter.

Some builders believe you can use a gearmotor with much less torque than we recommend. There is an extended discussion on gearmotors for lifting arms in this archive: search the page for 'RS-540'.

A: That's a fairly complete list -- you might want to remove 'lifter'. Active weapons add several layers of additional complexity onto an already complex undertaking. I recommend that first-time builders keep their robots very simple. You'll have enough new things to worry about with battery maintenance, R/C system set-up, traction issues, ESC mixing, driving, radio interference, wireing, tournament procedures, and repair problems.

A: The front wedge on the current #1 ranked heavyweight is mounted on a heavy rod that pivots in very sturdy mounts attached to the chassis. This allows the front of the wedge to pivot down and scrape along the surface of the arena, or drop and lie flat if the 'bot is inverted. The pivot point is low -- near the centerline of the robot -- which helps prevent the wedge from folding back under on heavy impact.

Zero-clearance wedges have a drawback: they can get hung up on joints in the arena floor. In a smooth arena a low-pivot hinged wedge can be a plus.

A: Mark J. here: it was in interesting experiment: a pneumatic powered rotary weapon with a claimed 150 horsepower. Record: 1 win, 1 loss. The problem was that it isn't possible to effectively store enough energy in the form of compressed gas to power such a weapon for more than a very few seconds. Pound for pound, electric batteries store much more energy than compressed gas cylinders.

A: That's way better than decent, assuming that you have a reasonable spin-up time.

Q: What size screws would be best for use as teeth on a 2.5" OD Beetleweight drum spinning at 18,000 RPM?

A: Mark J. here: unless that is a very narrow drum, it isn't reasonable to try to spin it at 18K RPM. Anything less than perfect dynamic balance will shake your 'bot around like jello in an earthquake. Even if it's perfectly balanced to start with, it won't be after the first hit. Worse, with two rows of teeth a tooth will pass by every 2 milliseconds -- you can't get any 'bite' into your opponent with teeth passing by that fast.

Slow down the drum by a factor of at least two. I'd suggest even slower. Speed and Joules are less important on a small drum than are bite and torque. For teeth try 3/16" hex cap screws. Anything smaller might shear off.

A: 'Crack Torch' never won a fight. Although they are crowd favorites, flame weapons are ineffective in any weight class. Your opponent would have to sit still while you positioned your flame weapon and held it long enough to do some damage. If your opponent is sitting that still, he's disabled and you've already won the match.

You can research the fight history of any robot at Botrank.com.

Q: What about Texas Heat? It is an effective lightweight flame thrower robot that is currently ranked 2nd. Wouldn't a similar design work for an antweight?

A: 'Texas Heat' is an effective wedge robot designed by the very experienced and well financed CM Robotics. It would be equally effective without the flame weapon. Flames are for show.

A: Small drums may have the rotating can of a brushless outrunner motor press-fit into the inner diameter of a custom machined drum. The motor mount becomes the support for one end of the drum. This design is simple and rugged, but gives you no control over maximum drum speed or spin-up time. I've also seen larger drum weapons with an internal friction drive running against the inner surface of the drum. In general, an external motor with a belt drive is a simpler and better choice.

A: 'Blendo' was a great robot and a real benchmark in robot design. Their Briggs & Stratton internal combustion engine was rated at 5 HP -- not much by today's standards for heavyweight spinner power.

A: There are several tips on 4-bar mechanism design in this archive. The T.i. Combat Robotics 4-Bar Simulator is a valuable design tool and a good place to start.

A: Combat robots are generally expensive. Larger combat robots are more expensive. You'd better have an overall budget of at least a couple grand to cover a minimal lightweight flipper.

Standard pneumatic components are off-the-shelf products available thru industrial suppliers. If you can put your system together from these standard components you can keep the price reasonable. Pricing for a budget CO₂ low-pressure (150 psi actuation pressure) lightweight pneumatic system might run something like this:

• an actuator cylinder: less than $100
• a useable 5-port valve: about $150
• a small paintball regulator: about $100
• a small CO2 paintball tank: less than $30
• assorted connectors, hoses, valves, guages, electronics, and mounts: another $100.

How much will it weigh? That's like asking how much a rock weighs -- it depends on the rock. A basic lightweight system might weigh as little as 6 or 8 pounds, but adding on performance will quickly add to your weight as well as the bulk of the system. Pneumatics take up a lot of room, so your 'bot may need to be larger which makes for more weight in the armor and chassis as well.

There is also the matter of assembling a safe and reliable system. Pneumatics are not for beginners. They can generate a lot of force and move very quickly. That is a recipe for severe injury if you don't know what you're doing. A first-time builder has plenty to worry with just getting the radio, electrical, and mechanical systems of a simple combat robot working correctly. Don't add on the complexity of an active weapon 'til you get the basics down.

A: Mark J. here: Although it is an interesting theoretical design, a flywheel powered flipper has a number of engineering challenges that render it inferior to pneumatic flippers. The idea is to store energy from an electric motor in a rotating disk, just like a conventional spinning weapon, but then to engage a clutch mechanism to transfer that stored energy via a linkage to a flipping arm or platform. The clutch must be able to engage quickly, transmit enormous force, and disengage quickly to prevent excess energy from ripping the weapon apart.

I've seen a few clever designs on paper, but no one has actually built a successful flywheel flipper. The complexity, weight, and expense would be much greater than a pneumatic system of similar capacity. Stick with pneumatics.

A: Mark J. here: HPA is High Pressure Air (78% nitrogen, 20% oxygen) right out of the atmosphere. HPA systems in combat robots typically have their tanks pressurized at 2500 psi or above -- likely much greater than the capability of your compressor. The RFL rules require HPA pressure tanks to be rated for al least 120% of the maximum pressure they will be used at, and to have a current hydro test certifying that capacity. High pressure paintball tanks are commonly used in small 'bots, and SCUBA tanks in larger 'bots.

Before you go any farther, read the Team Da Vinci Understanding Pneumatics page. That will answer a lot of questions you didn't even know you should be asking.

A: Mark J. here: there have been several attempts to build a successful invertible 'full body' spinner, with wheels sticking out both top and bottom. Ringmaster fought as a heavyweight at BattleBots 5.0 and had a good run, winning 4 fights before losing to Omega-13.

A rotating ring weapon is typically supported by a number of small wheels attached around the upper and lower perimeter of the chassis. The wheels run in grooves in the inner surface of the ring, supporting the ring and allowing it to rotate.

A: Mark J. here: it sounds like you're planning a small diameter weapon with the intent to grind away at the bottom armor rather than a traditional impact spinner weapon. What you're building is more of a mobile power tool than a sledge hammer. Joules of stored kinetic energy don't apply -- you're not going to rely on stored energy to do damage, but on continuous application of power.

You'll need enough power to keep the weapon spinning while in contact with the other 'bot and their full weight bearing down on the spinning blade/bar/grinder. I don't have any simple formulas to calculate the power you'll need for that. I'd suggest keeping the RPMs relatively low and gearing for torque.

A: Mark J. here: each of the power options you list have strengths and weaknesses:

• Pneumatics components are widely available and can provide high thrust, speed, and design flexibility. They do require additional support equipment and add complexity to the confined space inside your robot.

• Linear actuators are self-contained, compact, and have the same design flexibility found in pneumatic systems without the need for additional support equipment. They are, however, much slower and potentially less powerful than pneumatic systems.

• An electric gearbox capable of supplying the torque needed for a 12 kilo robot would be bulky and heavy. The T.i. Combat Robotics 4-Bar Simulator can give you torque requirements for a gearbox powered 4-bar lifter. You might be surprized by the torque required.

A similar amount of lift should work well with a pneumatic system. You may be able to get much more, but you did say 'lifter' and not 'flipper'.

A: A true 'press fit' involves machining the inside of the drum to a diameter a couple of thousandths of an inch smaller than the outside diameter of the motor and using a hydraulic press to shove the motor into place. If your machining isn't accurate, the drum won't run true.

A: For weapon power, two E-150s would weigh nearly twice as much as an S28-150 and produce 1/3 the power. Weight is the greatest challenge in building a combat robot. You need to make the best use of every ounce. The weight you save in the motor can go into the spinner disk for more stored energy and damage potential, while the added power will improve your spin-up time and maximum RPM. Well worth the added expense.

A: If you're building a large weapon I would recommend putting all available weight into the battery, motor, and weapon mass rather than on structures that do not increase the impact and energy of the weapon. The Robot Wars housebot 'Dead Metal' used a circular saw weapon that swung forward and down, but a cutting saw does not produce anywhere near the kick-back force of a rotary impact weapon.

Every action produces an equal and opposite reaction, so your weapon arm would need to absorb the same amount of force that the weapon inflicts on your opponent. That would take a massively reinforced arm assembly! Keep it simple.

A: The energy stored in a spinning weapon is dependent on the speed of the disk, the mass of the disk, and how the mass is distributed on the disk. There is also the question of the time needed to spin the weapon up to an effective energy level. Search this archive for information on calculating weapon energy, appropriate energy levels, and spin-up time.

Q: How were the chains attached to the weapon sprockets on 'The Judge'?

A: Careful examiniation of the photo at right shows that the last chain link is bent inward toward the shaft and riveted thru the sprocket. An animation of this weapon is in this archive -- scroll down about 40 questions.

A: Yep.

A: There are a lot of variables in calculating weapon battery requirements, but we gave a 'rule of thumb' and approximate formulas in an earlier post. Search the archive for 'battery capacity'.

Q: I took a look at your formulas to figure out weapon battery requirements and added them into your Spinning Weapon Excel Spreadsheet. It's formatted exactly like your work, and the formulas are all correct. Would you like to add it to your spreadsheet?

A: Nice idea! Michael Maples sent me his addition to the spreadsheet and it's way cool. I've replaced the downloadable file with the new version. Thanks, Michael.

A: Mark J. here: ideally you should be able to meet the full stall current demand of your weapon motor. Current equals torque, and if you can't supply the current the spin-up time of your weapon will suffer. From a practical standpoint, if you can supply 80% of your weapon motor stall current you'll be fine.

Note that LiPoly batteries can be damaged by current drain above their rated capacity. If your LiPoly pack is rated below stall current, you might want to use a servo slower module to 'feather in' the throttle to your weapon ESC and avoid the big start-up amp rush.

A: How about the classic DeWalt 18 volt drill motor? At 24 volts it kicks out 1.5 horsepower and it's less than a third the weight of the S28-150 magmotor. The problem with it and other small brushed motors is high RPM. You'll have to figure out a weapon drive that will stay together at 20K RPM.

Check out the performance of your weapon/motor combination with the Team Run Amok Spinning Weapon Excel Spreadsheet.

A: Mark J. here: 'Green Wave' was a middleweight spinner with the weapon motor mounted above the robot body on a short, stationary (dead) shaft. The motor support shaft also functioned as the axle upon which the bar weapon spun. The design allowed the bar weapon to be mounted very low on the chassis.

'Green Wave' had serious problems with weapon bearing disintegration caused by heavy off-axis loads on impact. To prevent this problem with your 'bot, the weapon bar assembly should be supported by two bearings separated on the shaft to spread the load and prevent twisting. I'd suggest one bearing in the belt drive pulley and a second bearing in a flange mount on the bottom of the weapon bar.

A: Eggbeaters take enormous punishment. Most are cut from a single plate of metal for greatest strength.

A: Mark J. here: come on, guys -- you've got to give me more info than the general type of weapon and the weight class! It's like asking, "What should I feed my yellow pet?" Dog? Canary? Goldfish? See comment in the next question below on 'how thick'.

A: Mark J. here: my standard answer to a 'how thick' question is, "as thick as you can make it and still make weight." In this case, a thicker drum wall will mean more energy storage and more durability. Like most design factors in a combat robot, there is a trade-off in adding weight to one system while pulling weight out of another.

A typical drum makes up around 20% of the total weight of the 'bot. About 3/16" aluminum might be a good starting point for a lightweight drum, but the size, speed, and design of the drum can modify this.

A: Optimum drum tooth length depends on tooth spacing and the axial speed of the teeth, not the size of the drum. You want your target to get as far inside the arc of the rotating teeth as possible in order to get maximum 'bite'. Faster spinning drums require wider tooth spacing and run shorter teeth, while slower spinning drums can have closer and deeper teeth.

You need to calculate the amount of time between teeth when the weapon is up to speed, then calculate how far your 'bot is likely to move forward in that time period when attacking. That distance is the optimum tooth length.

Alternately, since your attack speed is so variable, you could just stick the teeth out about 3/8" and go for it!

A: Bolts are not ideal drum teeth -- but they are inexpensive, easy to obtain, and simple to replace. The simple blunt impact is fairly effective, plus a little time with a hand file or Dremel tool can put a nice, sharp edge on a bolt head.

A: Mark J. here: the Team Whyachi TWM3R gearbox has a nominal 1 inch titanium shaft with a 3/8-16 tapped hole in the end and a .25 inch keyway down the side. This makes it convenient to lock a keyed weapon hub onto the shaft. I'm sure that Team Whyachi would be willing to machine a suitable weapon hub for you.

Note that a keyed hub has no slip or give and will transmit all weapon shock loads back to the gearbox, potentially destroying it. I prefer a slipable weapon drive like a v-belt, but the Team Whyachi gearbox is built to take high loads. Your choice.

A: Mark J. here: If you are using hardenable metal it's a waste not to harden them, but I see a lot of non-hardened drum teeth. For longest life and greatest damage, harden the teeth.

A: There is no maximum pressure specified in the RFL rules, but all components in the system must be rated for the pressure used. Above 250 psi, components must be over-rated to 120% of the pressure used. Systems operating above 2500 psi require pre-qualification by the event officials.

Consult section 7 of the RFL rules for details.

A: Son of Smashy used an electric winch to reset the spring after each strike. The difficult part in building a set-up like this is the release mechanism that allows the weapon to fire then re-engages the winch. In theory it could be done in any weight class but pneumatic systems are more powerful, quicker to reset, and can be built from off-the-shelf components. Spring powered weapons are rare for a reason.

Search the archive for 'winch' for other posts on this topic, and scroll down about eight questions for the post on 'Red Square'.

A: Axes are neither popular nor effective weapons, but I've seen hardware store axes, picks, hammers, and chisels used. High energy weapons usually have custom made impact pieces.

A: The Etek motors have very large armature intertia which will cause the rotor to shift on the shaft if stopped abruptly. This effectively destroys the motor. To prevent this you should drive a rotary weapon with a slipable system like a v-belt. You may need several parallel belts to handle the power output of twin Eteks. Gears, chains, or timing belts will lead to trouble.

Interestingly, the battery capacity needed for a rotary weapon is more dependent on the weight, geometry, and speed of the weapon than on the motor powering it. An example, with formulas, of calculating spinning weapon battery capacity requirement is given in a previous answer in the Radio and Electrical archive.

A: You can use a slice of well-frozen pepperoni pizza if you like -- the result will be much the same. Neither the pizza nor the knife are designed to take large loads at the pointy end. Either will simply shatter on good impact. You need something that you could hammer into concrete. Repeatedly.

A: KillerHurtz is an unusual robot in nearly all design areas: all-plastic chassis, combined differential and pivot steering, computer controlled ESCs, and a pneumatic actuated overhead axe weapon with a chain drive.

The weapon is powered by a 100 mm bore pneumatic cylinder operating at 150 psi. The cylinder is attached to a lever directly connected to a large chain sprocket. A chain carries the force to a smaller sprocket attached to the axe. There is a good photo on an archive of their site: KillerHurtz Design page.

Q: Did KillerHurtz use a double-acting pneumatic cylinder?

A: Yes. The pneumatic plumbing appears to be incomplete in the photo -- see the KillerHurtz pneumatics page for details.

A: Frostbite's thresher is simply eight short vertical bars arrayed along a common central support shaft. Mount the shaft on bearings and spin it with a belt drive.

Mark J. here: the most efficient design for a spinner weapon is a drum -- it can store more energy for size and weight than other configurations. A disc weapon is less efficient, and a rotating bar even less efficient. Multiple short bars, like the thresher, are very poor at storing energy.

A: Mark J. here: it takes just a little extra work. Input a double-thick bar of the correct length and width to represent the two side bars on your eggbeater. Manually calculate the mass of your two crossbars and represent them with a cylinder the diameter of your weapon with a thickness that brings it's mass equal to the calculated mass of your crossbars. That will get you very close to correct.

A: There is plenty of information on drum weapons in this archive. You may also want to get a copy of Combat Robot Weapons by Chris Hannold -- it has specific information on many types of weapons in detail that I cannot afford to provide here.

A: A leaf spring is a curved metal plate that has been tempered to 'spring back' to it's curved shape after being flattened. One end of the leaf spring is securely fastened to a chassis and the other end is winched down to flatten the spring, This stores a large amount of energy in a thin and flat profile. When another 'bot is on top of the spring the free end is released and the spring tosses the other 'bot skyward.

The principle is simple, but designing the winch and release mechanism is tricky. Team Whyachi's middleweight 'Red Square' was a successful leaf spring flipper -- ranked 15th, with a record of 19 wins and 12 losses.

Q: Would it be possible to make a spring powered flipper in a featherweight?

A: Possible, sure -- but there are reasons that spring flippers are not a popular design. Team Whyachi was successful because they are a very experienced team with great resources.

Q: How would I make a spring powered flipper in a hobbyweight?

A: No answer that I can give in a couple of paragraphs is going to help you much. Very few spring flippers have been built due to the difficulties in constructing a releaseable winch system to reset the spring. The mechanism from an automotive power seat would be a good start, but custom machine work would be needed for the release spool. I'd suggest building something else.

A: Mark J. here: Newtonian physics states, "For every action there is an equal and opposite reaction." When your spinning weapon strikes your opponent, equal forces are applied to both robots.

If the front edge of your weapon is spinning upward: the force applied to your opponent will be upward and your robot will be subject to an equal downward force. Since a solid surface is holding your robot up, nothing much will happen to it. Since nothing except gravity is holding your opponent down, they will fly end over end into the air -- greatly impressing the judges and spectators in your favor.

If the front edge of your weapon is spinning downward: the force applied to your opponent will be downward and your robot will be subject to an equal upward force. Since a solid surface is holding your opponent up, nothing much will happen to them. Since nothing except gravity is holding your robot down, you will fly end over end into the air -- greatly impressing the judges and spectators, but not in your favor.

Upward spinning, please.

A: Mark J. here: The 'best way' depends on a balance of your budget, experience, skills, and expectations.

• Pneumatic lifters are very quick and powerful but require complex, bulky components and a fair bit of weight allowance. Pneumatics are applicable to either a simple single-pivot lifter or a more efficient 4-bar design.

• Electric actuator lifters are typically slow but are comparatively compact, safe, and simple to support and maintain. Again, they can be used with single pivot or 4-bar lifters.

• Driving one arm of a 4-bar lifter with a gearmotor requires a heavy gearbox capable of handling enormous torque, and careful design work -- but the result can be quick and effective.

A: I can't find a picture of the weapon used by middleweight 'El Diablo' at BattleBots 5.0, but I did find some construction notes in a build report:

The weapon was made from four 1" thick polyethylene discs, each with a central bearing riding on a stationary steel shaft. Bridging across the four discs on opposite sides were two 1/4" thick 6160 aluminum plates with bolts threaded thru from the back side for 'teeth'. The entire assembly weighed 24 pounds and was reduction belt-driven by a Bosch GPA 750 motor to about 1300 RPM.

'El Diablo' went 0 for 1 at BB 5.0, losing their opening match to 'Ankle Biter'. Heavyweight 'El Diablo Grande' did better with a similar weapon, winning 3 matches before a loss to 'MechaVore'.

You can find pictures of El Diablo and El Diablo Grande as they fought at BB 5.0 at robotcombat.com, under 'Event Reports'.

Thanks! I'd forgotten about their extensive collection of BattleBots photos.

It looks like the team changed the polyethylene and aluminum design before the competition. Those appear to be four thick aluminum discs with steel striker hooks bolted in. I don't think I could recommend a similar design -- a drum weapon of the same size, weight, and speed would store much greater energy than the 4-disc setup.

A: Drum weapons are not projects for the average builder without access to and experience with precision machine tools. An error of a couple hundredths of an inch will leave you with a useless paperweight. A competent machine shop could certainly make a custom drum weapon for you, but the cost would be prohibitive (hundreds of dollars). Consider switching to an eggbeater weapon -- much easier to construct and balance. See also the previous post on balancing rotary weapons.

A: Clan MacCanIKill's 'Ziggy' has a 4-bar flipper mechanism with a pneumatic actuator acting on the front bar. Click the photo for a larger image. The pneumatics operate on high-pressure nitrogen at up to 3000 psi. Most pneumatic flippers use carbon dioxide gas which gives more 'flips' per tank but which also restricts their maximum pressure to about 850 psi.

In spite of the high-pressure nitrogen pneumatics, I don't believe that Ziggy is more powerful than other top-rank flippers. The 'laydown' position of Ziggy's actuator cannot match the mechanical efficiency of the twin upright actuators in Toro.

A: 6AL-4V titanium, about 5/16" thick should do.

A: Team Entropy never had large or complete site, and what they had was taken down about a year ago. I can direct you to an archived version of their site that has a picture of 'Redrum'.

Q: I once heard an announcer say that the new Redrum's internal combustion engine is located INSIDE the drum! Is that true? Is that even possible?

A: Yes, and yes. It's a very big drum. If you think about a drum as being kinda like a full body spinner on it's side, you might get more comfortable with the idea.

A: My advice: stay away from Internal Combustion Engines (ICE) in combat robots. They seem like such a good idea, but every team I've ever seen use an ICE has suffered multiple problems at tournaments:

• the engine becomes inexplicably difficult to start;
• the engine stalls at idle waiting for the match to start;
• the engine delivers one hit and quits;

A: Mark J. here: 'SlamJob' kinda cheated. Scott Kincaid built an ingenious system that used two pneumatic cylinders that worked in sequence to accelerate the weapon arm. Power is not evenly applied over the swing, but it worked very well. See if you can figure it out from this early construction photo of SlamJob. Hint: neither of the cylinders are physically attached to the axe arm, except for a retract cable.

`The Judge' did not cheat. Jacha Little built a system that evenly powers the hammer throughout a 180 degree swing. You can find descriptions and pictures at the Team Mechanicus Website (archived).

Q: I looked at the pneumatic system for the hammer on 'The Judge' on their website. It's kind of confusing to me. It looks like a rack and pinion but it's backwards, and has chains. Do you think you can explain it better to me?

A: You've correct, it is a reversed rack and pinion. The pneumatic cylinder pushes and pulls the 'rack' in a straight line. The 'pinion' is the pivot point of the hammer: it swings the hammer in an arc as the rack is moved.

Their 'rack' is actually a pair of heavy power chains, and their 'pinion' is a pair of chain sprockets. One chain is wrapped clockwise around the sprocket and attached, the other chain is wrapped counter clockwise around the other sprocket and attached. One chain is always under tension and can transmit power to the hammer. There is less chance of 'jumping a tooth' under high loads with the chain system.

Q: In the illustration, which side is the front end of 'The Judge'?

A: The layout in 'The Judge' has the pneumatic cylinder at the rear of the robot, so 'front' would be to the right.

Q: So, The Judge's weapon can be compared to a spring loaded weapon, in the sense that natural physical forces fire the weapon (spring) and a special mechanism is required to reset the weapon into a ready position (winch), right?

A: No: the pneumatic actuator powers the weapon in both directions. I've modified the illustration to better show the action. Hammer motion and force is roughly equivalent in either direction.

Q: How does 'The Judge' get so much power from retracting the cylinder?

A: 'The Judge' uses a 4" bore double-acting cylinder that produces (almost) equal power when extending or retracting. Gas pressure to the actuator is under microprocessor control to quickly vent the cylinder for quick cycling. Take a look at the Team Da Vinci pneumatics page for an explanation of double acting cylinders and for a different animation of the hammer on 'The Judge'.

A: Our heavyweight flipper has a very large CO² tank that will power the weapon thru multiple matches. We also designed the tank mounts and connections to allow quick removal and installation of the tank without tools. With careful design tank changes aren't a problem, and if you enjoy tossing your opponent in the air there really isn't an alternative.

Note: High Pressure Air (HPA) tanks can be refilled with the tank in the robot, but CO² tanks need to be 'pre-chilled' and filled with a specific weight of liquefied gas -- not easy to do with the tank in the robot. Check with your event organizer for specific pressure tank refill regulations at the event.

A: Many: spin-up time, optimum speed... the same elements that apply to larger rotary weapons apply to fairys. Read thru this section of the archive and you'll find plenty of pointers.

A: Go take a look at the Antweights - VDD Kits page at the Robot Marketplace for parts and complete saw weapon kits.

The idea of a spinner weapon is to store energy in a rotating mass and release a big burst of energy into a your opponent, sending them flying. if you make the weapon too 'small and light', it will just run into your opponent and stop. Browse this section of the archive for design ideas.

A: Not a great idea. See the previous post in the Ask Aaron Archive.

A. Sewer Snake is from the stable of Matt and Wendy Maxham's Team Plumb Crazy. It does better than just compete -- it was the 2005 RFL heavyweight champion, it is the #1 historic ranked heavyweight robot at botrank.com, and is a member of the Combat Robot Hall of Fame.

I wrote to Matt and Wendy for some details on Sewer Snake's weapon. They sent the following description and photos:

Q: Isn't Sewer Snake currently a superheavyweight?

A. Sewer Snake fought very briefly (and unsucessfully) as a superheavyweight in 2006. It returned to the heavyweight rankings to fight at the Combots Cup in 2007.

A. Drum weapons are usually cut from a section of thick-walled tube, or machined from a solid block of metal. Getting a rolled sheet of thick metal perfectly round and well balanced would be very tough. There is a lot of material about drum weapons on this archive page.

Q: Where would I find the tube I would need to make my drum?

A. It would help if I knew whether you were building an antweight or a superheavyweight. The Robot Marketplace metals department has a good selection of aluminum and steel tube in assorted sizes.

A: Mark J. here: something tells me you're not building a combat robot. An H-bridge provides on/off/forward/reverse control for a motor -- too crude for proper combat robot control. Our robots run either full ESC control for drive motors or relay H-bridge control for lifter weapons. We don't have any experience with selection of H-bridge IC chips.

Q: I'm using a resistor in series with the motor to lower the voltage when the current is big. The H-Bridge would also take a voltage drop, right?

A: No. Resistors are not voltage control devices -- they are current limiters. Adding a resistor in series on the motor circuit will limit the maximum current the circuit can pull, but because of the variable back EMF the motor generates in operation the voltage the motor 'sees' will depend on the load the motor is under. At low load conditions the voltage across the motor brushes (and in the rest of the circuit) will be very close to full available voltage. At high loads, both the voltage and current at the motor will be reduced by the series resistor. A series resistor in the motor circuit beyond the H-bridge will not impact the voltage at the H-bridge.

Q: I'm not interested in speed control but in torque/current/dutycycle and direction control, with a high current output (for the same FF-180 BaneBots motor). Would an ESC do the work for this purpose? If so, which do you think is a better choice, the H-bridge, or the ESC?

A: Yes, an ESC can control torque/current (they are equivalent in a permanent magnet DC motor) and provide directional control. The H-bridge and a series power resistor can perform similar functions. If I knew what you were actually using the motor and controller for I might be able to give an opinion on which is 'better'.

Q: Remember the questions about a worm gear driven robotic gripper I posted before? I want to control the pinch force of the gripper and of course it must be able to use the full power of the BaneBots motor. Since I also need to control direction, I thought of an H-Bridge like the LMD18200T, but its operating voltage is from 12 to 55V.

A: Yes, welcome back! Go take a look at Pololu Robotics line of motor controllers. I'm sure you can find something there that will suit your application. In particular, the 'MC33887 Motor Driver Carrier' looks like a winner, but there are several other options. Add a series power resitor to the motor and you'll be in great shape. Or you could go with an ESC...

Q: If I were to use an ESC for this purpose, can you tell me how to connect the ESC to control torque/current instead of speed?

A: The term 'Electronic Speed Controller' is not really accurate: the device controls the power available to the motor rather than directly controlling speed. It effectively controls amperage, which directly controls torque and, only indirectly, speed.

Q: OK, so the signal I need to send to the ESC would be the same as the signal needed to control a servo, right? A 'high' (4 to 6 volt) pulse on a 20 millisecond period where:

• 1.0 to 1.5 millisecond pulse is proportional high to low power 'forward' rotation;
• 1.5 millisecond pulse is neutral - no rotation; and
• 1.5 to 2.0 millisecond pulse is proportional low to high power 'reverse' rotaton.

A: You've got it exactly correct -- the same signal you sent to the servo you're replacing with the gearmotor. You can control torque by limiting the pulse width to a range nearer the 1.5 millisecond 'neutral' range.

Note that 'forward' and 'reverse' are relative -- you can always reverse the motor leads if the direction is not correct for your application.

Q: What ESC would you recommend - perhaps the BaneBots 3 Amp continuous / 9 Amp peak ESC?

A: The Banebots ESC should work very well for you, It is small, thermally protected, reversible, and has a fail-safe startup routine.

A: Team Raptor's successful series of lightweight robots (Alpha / Beta / Gamma Raptor) all used 4-bar electric lifters. I believe the short rear arm of the lifter was powered by a high-reduction gearmotor -- possibly an Astro Flight cobalt.

Team Raptor's website is no longer on-line, but you can access an archive of their site from 2005.

A: Mark J. here: electric motors can power effective lifters, but a flipper must unleash enormous power within a very short time period -- no practical electric motor can deliver that much instantaneous power. You could use an electric motor to spin up a big flywheel and then release the stored energy into a flipper, but the weight and complexity of the system would be prohibitive. Pneumatic flippers are cheaper, lighter, simpler, and more effective.

Q: I didn't necessarily mean "flipper". Is it possible to create a platform that rises up and forward (like Firestorm from the UK Robot Wars) without using a pneumatic cylinder or linear actuator of any kind?

A: Sure! It wouldn't be as fast or powerful as a pneumatic system, but it can certainly work. Driving the platform from a shaft at the hinge point requires a lot of torque and a heavy duty gearbox, but other designs -- like a modified 4-bar linkage (animation at right) -- can reduce the required torque and simplify construction and control elements.

A: BioHazard's lifter uses a pair of highly modified 12 volt linear actuators running at 24 volts. Estimated power from each of the twin actuator motors is over 1 horsepower, with each actuator delivering 1400 pounds of thrust. The motors used are not standard robotics items. More information on the actuators can be found at the BioHazard website.

A: A price tag of nearly $1000 is certainly one reason. The Perm PMG 132 motor can crank out 34+ horsepower at 72 volts and would make a mighty weapon motor for a heavy or superheavy 'bot. However, that much power requires a big stack of battery capacity and a very serious motor controller. Overall, you might be better off with an internal combustion powerplant for a mega-weapon drive.

A: I don't think you've read closely enough. Most of Team Run Amok's active weapons, from the lifter in our beetleweight champion 'Zpatula' to the Bosch-powered spinner on our heavyweight Robot Wars competitor 'Run Away', are controlled by relays. We don't have anything against them -- in the right application.

The choice of a mechanical or electronic relay versus an ESC for weapon control depends on many factors. For very large weapons, a reliable relay or contactor may cost as much as an ESC and will certainly weigh more. The high amperage surge that comes along with using a relay is hard on the batteries, and for a very large weapon the capacitor you suggest would have to be huge to be any use at all against a multi-hundred ampere spike.

As far as combat robots being 'doomed to be ripped apart', we have never lost any electronics to opponent caused damage. Maybe we're lucky, or maybe we just build 'em right.

A: Yes. The BaneBots 125:1 36mm planetary RS-540 gearmotor could be used on a 12" lifting arm for a hobbyweight, if the shaft was well supported. It should provide over 20 pounds of stall force at the end of the arm at 12 volts. You might also consider the 256:1 42 mm planetary RS-550 gearmotor for it's larger diameter shaft and greater torque.

Q: You told the guy that asked about the RS540's on a lifter to consider the 256:1 RS-550 motor. I dont know about the 42mm gearbox, but the 38mm 256:1 gearbox's durability is very poor. You may want to say something to that guy before he gets the 256:1 and it busts on the first attempt. Maybe recommend the 64:1 since it has way more than enough torque, or preferably the 48:1, again way more than enough torque.

A: The BaneBots 256:1 38mm gearbox is a 4-stage unit and is not offered with the RS-540 or RS-550 motors. The 125:1 RS-540 38mm gearmotor is a stronger 3-stage design -- the same as the 64:1 and 48:1 boxes -- and should be fine in the hobbyweight application I recommended. The 256:1 42mm gearbox is much stronger than the 38mm gearbox and should also be fine for a small lifter.

I don't agree that the 64:1 and 48:1 gearboxes with an RS-540 motor would provide 'way more than enough torque'. For reliability and maximum lift speed, you want to load the system to no more than 60% of stall torque at maximum lift weight. With a 12" arm and allowing for gearing losses, the RS-540 motor with a 64:1 box stalls at about 10 pounds, and the 48:1 box stalls at less than 8 pounds.

Q: Unless you are fighting a very tall, thin bot all 12 lbs will never be on the lifting appendage at one time, unless some is higher than the point, like say a 6" wide bot is on the arm, with one edge at the tip and extending towards the shaft, and in that case it wouldn't be stall torque / 12", it would be stall torque / 6".

A: The needed torque is calculated by measuring from the pivot point to the location of the center of mass of the opponent -- not to the closest edge. In your example, the center of mass would be 9" from the pivot. If your arm tip is under the center of mass, you'll need full-weight lift capacity regardless of the width of the oponent.

Having the full weight of the opponent on the end of your lifter is not as unusual as you think. The end of the arm can and does lodge in 'bot openings, and large lift torque is needed when your opponent is pinned against the arena wall.

Q: No one designs their lifters with that much torque. It's a waste of weight and generally unrealistic unless you have a super huge lifting spatula or whatever at the end of the lifting arm. Biohazard is the greatest heavyweight lifter of all time, and it couldn't lift 220 pounds out at the end of it's arm. I'm not trying to criticize your answer, just trying to help another bot builder.

A: Mark J. here: we design our lifters to lift the full weight of the class they enter, and then some. Other builder do as well. For example, Team Cosmos' very successful hobbyweight 'IO' has a minimum 15 pound lift capacity. Carlo Bertocchini claims that Biohazard will lift 220 pounds, and the photo at right appears to show it doing just that.

Team Run Amok's lifters do have large surfaces that focus the lift toward the end of the arm, but the original questioner did not specify a design or even a weight class.

As for being a waste of weight, the BaneBots 125:1 36mm gearbox weighs the same as their 64:1 and 48:1 boxes. With the higher gear reduction you get greater lift capacity, less motor stress, lower amperage draw, and faster lift rate at heavy loadings. I'll back-up Aaron's recommendations.

32.32 in this case represents G, the gravitational constant. This isnt the kind of question you normally answer, but is my math correct? I know that the answer probably won't be realistic, but it would be fun to have a stat like that.

A: Mark J. here: You're not quite ready for the physics final exam. Joules don't directly convert to distance, but can convert to height. You're also mixing english and metric units -- keep the height in meters, mass in kilos, and the gravitational constant in metric units (9.8, not 32.3). The equation itself is correct, but can be further simplified:

So, a weapon with 6000 joules of energy has the potential to toss a 100 kilogram opponent: 6000 / (100 * 9.8) = 6.1 meters high. You're correct in saying the number will not be realistic, since it assumes a conversion of all of the weapon energy directly into vertical motion of your opponent.

Q: Wow! That formula yields 18.7 meters of potential height for a hobbyweight with only 1 Kj of energy. That really makes me appreciate how much power that is.

To calculate distance, I figure that an object may get thrown at about a 45 degree angle from a drum hit, which allows the trajectory to be modeled with height equal to distance -- a trajectory ratio of 1:1 with the height and distance both half of the original height: ( 1000 joules / ( 5.45 kilos * 9.8 ) ) / 2 = 9.4 meters high and 9.4 meters far. Is that about right?

A: You're more than a little off on your calculation of trajectory. With a 45 degree launch angle, the energy will be equally divided between vertical and horizontal vectors. The veritcal speed slows, stops, and then reverses under gravitational acceleration. However, the horizontal speed of an object thrown on a ballistic path stays constant (negating drag). As a result, an object thrown with identical initial vertical and horizontal speed (45 degree angle) will travel four times further than the maximum height it reaches. The formulas are:

Your maximum height calculation is correct, but the theoretical maximum distance thrown would be 37.4 meters. There is a very nice trajectory calculator about half-way down the page at the HyperPhysics Trajectory webpage.

Q: For every action there is an equal and opposite reaction, so won't half of the energy of the weapon hit go to throw my 'bot backward a distance equal to the distance my opponent is thrown forward?

A: The vertical component of the hit energy will act against the arena surface, resulting in no movement of your 'bot and full transfer of that energy into the height of your opponent's trajectory. The horizontal component of the hit energy will act equally on both 'bots, but your 'bot will be sliding across the arena surface with high friction resistance, while your opponent will be flying thru the air with minimal resitance. You'll get a little 'kick back' distance, but it will not be anywhere close to the distance your opponent will fly.

There are lots of factors that will reduce the actual distance the opponent will travel, but remember that this topic started out as just a 'brag number' -- a maximized ideal just for shock and awe.

A: Mark J. here: MOI is a measure of the resistance of an object to a change in rotation. It is dependent on the mass and shape of the object, and the axis of rotation. If you want to do it the hard way:

• Flat rectangular bar, axis thru midpoint: MOI = 1/12 Mass * ( Width² + Length²) -- yes, that is 1/12th, not 1/2.
• Disk, spinning on symmetry axis: MOI = 1/2 Mass * Radius²
• Tube, spinning on symmetry axis: MOI = 1/2 Mass * ( Inside Radius² + Outside Radius²)

The easy way is to use the Team Run Amok Spinning Weapon Excel Spreadsheet and get the MOI plus the spin-up time for your weapon as a bonus.

Q: OK, I got the MOI, but how do I calculate spin up time?

A: See the previous article on kinetic energy and weapon spin-up time for a link to a Paul Hills' site with all the math. The Team Run Amok Spinning Weapon Excel Spreadsheet will calculate the spin-up time for you.

Q: How many joules of energy is average for a spinning mass weapon on a hobbyweight?

A: See the previous article on kinetic energy and weapon spin-up time.

Q: My hobbyweight has a drum with an MOI of 21 lb-in². It's being driven by a 24V DeWalt motor at 24V. How much capacity should the battery for the DeWalt have?

A: Energy required for spin-up is dependent on the gear reduction you choose for the drum. I'm guessing you'll want about a 3:1 reduction to keep the drum RPM reasonable. That'll spin it up to 1000 joules in 1.8 seconds. That's gonna be some weapon!

There are too many variables to calculate a 'hard number', but I'd estimate that a high-amp 600 mAH pack would give you enough energy for a three-minute match. Common practice is to use a single battery for weapon and drive motors. See the article on battery selection.

A: Mark J. here: all of the blue input boxes (material density plus all dimensions) must be filled in for a bar, disk, or tube before mass and MOI can be calculated. If that isn't the problem, your spreadsheet may have been corrupted -- download a new copy.

The differing shapes, placement, and number of teeth on a drum spinner made it problematical to include a simple calculation box in the spreadsheet for them, but if you get a little creative you can approximate their presence. Calculate the total mass of the teeth, then increase both the outer radius and thickness of the tube to add in the extra mass. Alternately, for just a few teeth, increase the radius of the endcap(s) to bring the mass up.

A: How quickly you need to spin up depends on:

• The size of the arena;
• The ability of your 'bot to dodge the first attack; and
• How much energy you have in your weapon.

A: See the previous post on pneumatics and 4-bar lifters. The solenoid valve systems are electrically operated and connect to your radio system via a R/C Switch Interface.

Q: Where can I buy parts for a pneumatic hobbyweight flipper?

A: Hobbyweight and larger combat robots use standard industrial pneumatic components.

• For actuator cylinders, valves, tubing, and connectors, look in your phone book under 'Hydraulic Equipment and Supplies'.
• On-line try www.mcmastercarr.com. -- search for 'pneumatic cylinders' to get started.
• For small pressure tanks and regulators, check with your local or on-line paintball gear stores.

A: Mark J. here: The questions are getting tougher. A full answer to your question would make a good chapter in an advanced physics text.

Simplest case - where the flipper is positioned directly below the opponent's center of mass, and the flipper force perfectly vertical and uniform for the entire stroke length:

Complex cases - if you want to calculate the height achieved from off-center flipper hits, non-constant flipper force, and non-linear flipper vectors, break out your calculus text. You'll need to calculate - amongst other things - the moment of inertia for a specific opponent and axis of rotation. That ain't easy, and life's too short. For more information on Newtonian mechanics, visit the Hyperphysics website.

A: Search the Ask Aaron Archive for 'winch' to see previous answers. There's a reason you don't see many spring-powered weapons on combat robots: pneumatic systems are more powerful, faster, safer, and easier to build.

A: Take a look at Team Boomer's Fright Knight for a good example. Note that their weapon shaft is supported both above and below the plane of the disc. The supports form a triangular shape and are anchored directly and firmly to the chassis. The shaft anchor point is heavily gusseted with flat metal plate to spread the load to the supports. The support arms are also no longer than they need to be.

Q: Well, that type of design works well for 'Fright Knight', but I was talking about a 48 inch bar similar to 'Hazard' in design. I needed to know: how do I keep it from simply pulling out of the bearing at the base of the robot?

A: Sorry -- my mind reading is a little weak. You'll still want to support the shaft at two locations, above and below the drive pulley. If the pulley is pinned to the weapon shaft and tubular spacers are inserted as needed, the pulley will locate the shaft.

A: The hammer system is actually operated by chains and sprockets that emulate a rack and pinion. You can find descriptions and pictures at the Team Mechanicus Website (archived).

A: One... two... three... hey, it only has three bars! How come nobody ever counted before? The base counts as the fourth 'bar'.

A: Read the previous post about Breaker Box. The problem isn't power, it's finding a gearbox that can handle the huge amount of torque. A single RS-540 could provide enough power for a quick lift of 200 plus pounds at a gear reduction of around 1600:1, but the gearbox would have to handle 80,000 in-oz of torque!

A: You're going to transmit more torque than you probably think, but yes you can get away without bearings. Keep the axle size fairly large and lubricate with a little grease. Better still, use oilite bushings on the axles -- cheap and light.

A: Sure, but support the weapon shaft on both sides of your saw with bearings or you'll destroy the gearbox. Try the 5:1 ratio box for a small spinning weapon.

A: Electric lifters require great gobs of torque. See the earlier articles on electric lifters and 4-bar lifters. An electric hammer has pretty much the same trouble -- it's very difficult to get enough power to make one effective. Pneumatics work much better.

A: Mark J. here: yes, designing a 4-bar lifter system is 'sticky'. I recommend giving the T.i. Combat Robotics 4-Bar Simulator a try. It gives charts of torque and lift, as well as lift height and lateral displacement.

Our beetleweight lifter Zpatula uses a gearbox to power the front arm of the 4-bar lifter. The bar lengths between pivot points) are:

• Front: 3.375"
• Rear: 2.25"
• Top: 4.5"
• Bottom: 5.625"

Q: Ok, I got my measurements. How much weight should it be able to lift? I was thinking 30lbs would be enough for the featherweight sportsman class, since it would hardly ever experience the full weight of the opponent. About how high should it be able to lift?

A: I like your reasoning on the weight the lifter should be able to lift, but remember that the 4-bar simulator gives you the torque needed to HOLD a weight at a given position -- you'll need excess torque to be able to LIFT that weight. But you're right when you say the lifter will only rarely see full opponent weight.

How high is a tricky question that depends a bit on your attack strategy. The usual lifter attack involves breaking the opponent's traction and then shoving them into a wall, rather than trying to turn them over while everybody just stands still. I think 12 or 14 inches would be plenty.

The question you aren't asking is: how fast should it lift? Speed will depend on the weight on the lifter, and faster is certainly better. If you're set-up to 'hold' 30 pounds at stall, then a 15 pound load will take your lifter motor down to about half the no-load speed. I'd aim for no slower than 2 seconds to full lift at half load.

Q: One last question then I swear I'll leave you alone about the 4-bar lifter ; - ) According to the T.i. 4-bar calculator, my design needs a gearmotor with 1000 in-lb of torque to lift 30lbs of bot. That's a lot in a small enough package for a featherweight, so how do I calculate the thrust requirements for a linear actuator actuated 4-bar lifter?

A: You've discovered why there are so few electric powered lifters. Powering the shorter rear bar of the 4-bar assembly or shortening the overall length of the lifter can reduce the torque requirement, but it's still big.

Powering the assembly with a linear actuator (like BioHazard does) trades off torque for thrust, but the total power requirement remains the same. At the point of travel that calls for peak torque, you're still gonna need 1000 pounds of thrust at a right angle to a 1 inch lever arm, or 500 pounds of thrust on a 2 inch lever arm, or ( 1000 / X ) pounds of thrust on a 'X' inch lever arm. See the photo of BioHazard's twin linear actuator set-up to see how Carlo Bertocchini does it.

Don't forget that your chassis has to put up with the force of all that thrust as well. The actuator will push just as hard on the chassis mount as it does on the lifter assembly. Kinda makes you appreciate Carlo's design, eh?

A: Mark J. here: servos do not generally have a thermal shut-down. They are often used in applications where a shut-down could be more disastrous than risking damage to the servo! The HSR-5995TG pulls more than 5 amps at stall, which will quickly heat up the power controller and motor.

Failure mode will likely be abrupt rather than gradual -- a puff of smoke and zero response. You can reduce the heat build-up and stress by running the servo at lower voltage, but this will also reduce the available torque and speed of the servo.

Grippers in 'real world' robots generally use a worm drive gear train rather than the spur gears used in your servo. A worm drive does not usually require motor power input to hold against back-force on the output, so the drive motor can be powered down and the gripper will still hold.

Q: Do you know of some gearmotor with torque-size-power comparable to that of the HSR-5995TG servo with worm gear train included?

A: Nothing that powerful and that small. Tamiya makes several small motor/gearbox combinations that include a worm-drive stage, but their output is considerably less than your servo. You might investigate worm-drive mechanisms for automotive power antennas. They are larger, but have a cable output that would allow flexibility in positioning the gearmotor itself.

Q: Do you know someplace to buy a miniature worm gear train?

A: Try the Stock Drive Products/Sterling Instruments site. They have a large selection of small worms and wheels.

A: Mark J. here: the efficiency of gearboxes depends on their design, but forget about the gearing losses -- trying to put 476 foot-pounds of torque thru a BaneBots gearbox would instantly turn it into scrap. You'll need a gearbox rated for that type of torque -- search industrial suppliers like McMaster-Carr or Grainger, or look at truck/ATV winches.

Q: How much torque at the axle of that 20 inch long arm do you think I should have, since it will be a middle weight and the lifting attachment will weigh a lot?

A: Your torque calculations are about correct -- you'll want to aim for about twice as much lift as the load you expect to have on your lifter. Does your lifter arm really have to be that long? The problem is getting a gearbox heavy enough to handle that amount of torque. Why don't you write to Jim Smentowski and see if he'll tell you what gearboxes he uses on Breaker Box?

Q: OK. I talked to Jim and I think I should use an NPC motor because of their low rpm and high torque. McMaster-Carr also has worm gear reducers up to 60:1, but I still need to know how the gearbox efficiency will impact my torque output. How do I calculate the real output of my motor through a gearbox?

A: You will lose some power with any type of gear reduction. The general equation is: output power = input power * gearbox efficiency

Gearbox efficiency depends on many factors:

• the type of gears used (spur, bevel, crown, worm);
• the number of gear meshes under load;
• the number and type of internal bearings;
• the gear clearances in the design;
• the lubricant used; and
• the design and construction of the shaft seals.

This is a good time to remind builders that the best way to figure out if something works is to look at similar designs built by other builders. If you're building something entirely new, you're kinda on your own. Best luck!

A: Torque reaction overhead blades/spikes are legal and easy enough to build, but they would not be effective under the current damage/aggression scoring system.

For an explanation of the operating principle, see the Toecrusher website.

A: Mark J. here: `Toe-Crusher' uses a torque-reaction hammer weapon - a very simple weapon based on applying the torque of acceleration and braking to swing an overhead axe. Have you ever seen a motorcycle do a reverse-wheelie? The rider hits the front brake hard enough that the rear wheel comes up off the ground and the bike balances on the front wheel. If the rider braked even harder, the entire bike would flip over and crash into the ground in front of the braking wheel. If the motorcycle accelerates very hard, the opposite happens and the front wheel lifts off the ground.

A torque-reaction hammer works the same way. The robot is built with the weight balanced on the axle and a long arm with a hammer or pick on the end. When it accelerates, the hammer `pops a wheelie' and is thrown over to the rear of the `bot. When it decelerates, the hammer does a reverse-wheelie and is thrown over to the front of the `bot. Later versions of 'Toe-Crusher' had a wedge attachment stayed in contact with the floor by pivoting on bearings at the axles. Although showy, you don't get a lot of force out of this type of weapon and it is difficult to drive effectively.

A: Mark J. here: you'll need to know the stall torque and top speed of your electric motor, the shape and dimensions of the spinning weapon, the type of material you'll use to make the weapon, and the gear reduction between the motor and the weapon. The math gets a little sticky, but I wrote an Excel spreadsheet that will do the calculations for you:

As a rule of thumb, you'll want a minimum of 20 joules of energy per pound of weight class, and you'll want to spin-up to 10 joules per pound within 2 seconds. A modern high-energy spinner might have more than ten times that energy! You might have a little more spin-up time in a big arena, or if your 'bot is nimble enough to protect the weapon from premature impact. Smaller arenas used by sub-lightweight 'bots will require quicker spin-ups.

If you want to wade thru the math yourself, go to Paul Hills' page on Spinning Disk Weapons

A: Sorry, no. The hobbyweight class is dominated by passives (wedges, rammers, dustpans), spinning weapons (mostly drums/eggbeaters), and a few lifters. Spinning weapons store power delivered by their motors over several seconds of spin-up time as kinetic energy. A hammer that gains power over a fraction of a second from a similar motor just can't put together enough energy to do any damage.

A: The design of 4-bar systems is very technical and I've never found a good primer on the web. There is a 'Four Bar Front Bar' calculator that could be helpful if you were building an electric lifter.

You can get a very good education in general robot pneumatic systems at the Team Da Vinci Robotics pneumatics page. Remember: high-pressure pneumatic systems are potentially VERY dangerous. Don't get in over your head.

A: Yes, but it has several drawbacks and may not be considered legal under current RFL rules. Not everyone interprets the rules the same way -- check with your event organizer before you build!

You'd need to add a pressure rated 'T' fitting between the tank and the system shutoff valve before the pressure regulator. A second shutoff valve on the other branch of the 'T' connects to a male quick-disconnect refill fitting.

This adds weight, bulk, and complexity to the system. I think it's easier to design the CO2 tank mounts for easy release. You can then carry the tank (instead of the whole robot) over to the refilling station.

A: 'SJ' (it used to be called 'Slam Job') has a classic overhead pneumatic pickaxe -- two pneumatic cylinders acting together on a pivoting arm. Getting a pneumatic system to work over a long weapon arc is tricky. SJ uses one short cylinder to give the weapon an initial 'boost' at an optimum angle until the working angle for the second longer cylinder improves.

A: You'll need a narrow two-wheel high-speed drive and a long sturdy boom with a sharp pick or blunt heavy weapon on the end. The rest of the 'bot is much like any other 'bot. Search the archives for 'thwackbot' for more tips.

Thwackbots are not currently popular. The RFL Judging Guidelines demand constant aggression to score well and a simple thwackbot cannot spin and move toward an opponent at the same time.

A: Drum weapons are most effective when the teeth are spaced far enough around the drum circumference to allow the other 'bot to get close enough to the drum for the teeth to dig in and get some 'bite'. Don't put on too many teeth!

I don't know if big screws are better than custom teeth, but they're inexpensive and easy to replace! Make sure you have plenty of material depth in the drum to support the threaded shaft of the screw.

A: Mark J. here: It would help if I knew what size 'bot you're building. When in doubt, go bigger.

Screws and bolts are designed to take tension stress along their axis and are not ideal for enduring the shear stress that will be imposed when you use them on a spinning drum weapon. Their advantage is that they are inexpensive and easy to replace. Use a coarse thread -- it will be easier to remove when damaged.

A: Mark J. here: no -- weapon RPM is limited only by bearing friction, transmission loss, and aerodynamic drag. A very heavy weapon will eventually spin-up to the same RPM as a lighter weapon. However, given the information you list, you can calculate the approximate RPM of the weapon after a given number of seconds spin-up time:

If you'd like to learn more about the physics of kinetic energy weapons, read thru Paul Hills' page on Spinning Disk Weapons.

A: Mark J. here: the simple answer is 'all you can manage'. Successful FBS 'bots typically spin a shell representing around 20% of the total robot mass at speeds around 3000 RPM. I'll let you do the math to calculate the KE of such a weapon. The Team Run Amok Spinning Weapon Excel Spreadsheet will help.

A: Mark J. here: yes, milling a Full Body Spinner shell from one big chunk of metal does waste material, but I've seen it done. More commonly, somebody finds a short section of large diameter pipe at a scrap yard and welds a top onto it. I've also seen a tube formed from thick plate with a set of forming rollers, seam welded, and a top added. The key is precision to keep the finished product well balanced.

A: Flame weapons are allowed at the discretion of the event organizer. Check with the specific event you plan to enter for eligibility and special rules.

Note that flamethrowers are not effective robot weapons. They are for very experienced builders who just want to show off. If you have to ask how to build one, you aren't experienced enough to attempt it. I'm not going to encourage anyone to try dangerous construction beyond their skill level -- build something else.

A: Interesting question! I figure anything more than 45 degrees isn't a wedge, it's a sloped 'brick'. How much lower than 45 degrees you go is a function of how the wedge is going to be used. A very shallow angle is useful for defense and sneaking in under spinning weapons. A steeper angle is better for ramming attacks. A curved 'scoop' can be effective both for offense and as a 'spinner killer' -- I'd say that's ideal.

Q: What material would you use the create the "curved scoop"?

A: Titanium would be great, or steel if you have enough weight allowance.

A: Mark J. here: The more precisely a weapon is made, the better the balance is likely to be. If the weapon was hand drilled and cut, it isn't going to be anywhere close. Assuming the mounting bearings spin freely, you can 'static balance' the weapon:

• Remove the connection to the weapon motor (belt, chain, friction drive...) so that the weapon spins free.
• Position the robot so the weapon shaft is horizontal (parallel to the floor) and the weapon has room to spin.
• Give the weapon a spin with your hand and let it coast to a stop.
• Mark the part of the weapon that is at the bottom, closest to the floor.
• Repeat the spinning and marking about a dozen times. The part of the weapon that has the most marks on it is too heavy.
• Remove some material from the heavy part of the weapon, or add some weight to the opposite (light) side.
• Repeat the spinning, marking, and weight adjustment until there is not an obvious heavy spot.

A: Sure! Get hold of a copy of Robot Combat: Weapons by Chris Hannold, or Kickin' 'Bot by Grant Imahara. They both cover basic weapon construction materials, design, and technique.

A: Very ambitious plans! I would suggest starting with something simpler for your first 'bot. Exotic weapons won't do a lot of good until your basic mechanical systems are up to the task.

The current top-bots have heavy, high-speed rotating 'kinetic energy' weapons that hit like a speeding truck. The preferred armor in the heavyweight class is 1/2" thick titanium. Get to a tournament and see what the competition looks like before you decide on a design.

A: The first step is to carefully read the Robot Fighting League rules. The RFL rules are used by most robot competitions in the US. That will answer your questions about nail guns (not legal) and huge wheels (legal, but you're not gonna get thru 'bot armor that easy). Search the archive for suggestions on books for robot construction.

A: Nice choice! Your gear reduction will depend on the size of your undercutter blade. Take a look at Johnson Junior -- a S-280 powered undercutter with a big 8.5" blade and a gearbox. They're running a 5:1 reduction which seems about right for a blade that big. A smaller and lighter blade would use less reduction -- maybe 3:1 for a 5" blade. You can use the Team Run Amok Spinning Weapon Excel Spreadsheet to pick the best gear ratio.

Most builders like belt drives over gear drives for weapons because they require less precision, can absorb impact shock loads, and won't 'jam' as easily. A well designed gearbox can still be effective and rugged.

A: Brushless hobby motors are mostly sold to model aircraft builders. On a model airplane a 'firewall' is the flat panel at the front of the plane where the motor is attached.

Robot weapons put large side-loading forces on the motor shaft. Support the motor by mounting it as close to the weapon drive as possible -- 'behind the firewall' style.

Mark J. here: Engines in early airplanes were prone to catch fire. The 'firewall' was the solid bulkhead behind the engine that prevented flames from entering the cockpit. In that context it wouldn't make much sense to mount an engine behind the firewall, but the term survives.

A: For an antweight, that's probably too thick. For a heavyweight, that's probably too thin.

A: Team Inevitable Destruction's 'Cadaver' has simply hard-mounted the can of a brushless outrunner motor inside one end of a drum and put a support bearing in the other end. The force of impact is taken directly by the motor bearings, which is hard on the motor. A belt drive would isolate the motor from impact forces, give a faster spin-up, and would keep the weapon speed down enough to 'grab and toss' an opponent.

A: You think lifters are easy? I guess you can make an antweight lifter out of a servo without much trouble, but a bigger 4-bar linkage lifter is not easy.

A better question might be, "What is the simplest weapon that is effective at winning matches?" The answer might surprise you. My dad looked at the results from 20 recent robot tournaments to see what type of weapons did best. Take a look at the results: What Weapons Win?

A: A drum is just a large diameter metal tube with end plates and a few teeth welded to the outside. It's empty inside except for the shaft. The weapon shaft needs strong support, like any spinning weapon. Spin it with a belt drive from your weapon motor.

A: It's like the difference between a rotating bar and a disk. An eggbeater (like Team Sawzall's 'Switchblade') can't store as much kinetic energy as a drum that weighs the same -- the weapon has less rotational inertia. They're still effective, and more durable than a thin-walled drum.

A: Same as for any other rotating weapon: solid supports, correct RPM, protected drive, good balance, and quick spin-up. Drums are particularly hard to balance, so use extra care in precision construction.

A: The weapon/pulley shaft must be supported so that it remains aligned and does not wobble. This requires two bearing supports at separate locations along the shaft on either side of the pulley and/or the weapon. You'll need to adapt or fabricate bearing mounts strong enough to both support the correct pulley alignment and absorb the weapon loading.

Take a look at Team Basenji's antweight 'Bitsy Blade' at the Robot Riots 5 photos page for an example of a belt driven antweight spinner.

A: Technical question, Mark J. here: Briefly, gear drives require precision alignment and spacing to function properly. They can operate at very high RPM, but are not terribly efficient at transmitting power. Belt and pulley systems can tolerate some misalignment and are capable of absorbing the sudden shock loading a rotary weapon can produce.

Timing belts have small raised teeth along the inner surface of a flat belt that mesh with grooves in the pulley to reduce slippage and increase power capacity. They are very efficient at transmitting power and can operate at higher speed than regular belts.

I like belt drives for rotational weapon systems like bars, disks, and drums. The ratio of the number of grooves on the larger pulley to the number of grooves on the smaller pulley is your gear ratio. You'll probably want to try something around 4 to 1 for a large bar weapon.

Robotcombat.com has a selection of small timing belts and pulleys suitable for antweights in their in their Mechanical and Drive Components section. Their timing belt page has a calculator to help with correct belt length selection. Tower Hobbies has inexpensive, ready-made gearboxes designed for model airplanes that would be useful for an ant weapon. Search for `gearbox' at their site.

A: Wow - another technical question - Mark J. here: Power is calculated as the product of torque and RPM, so the real answer to your question is that both are equally important. A spinning weapon stores power in the rotating mass of the weapon and unloads it destructively onto your opponent. You need as much spinning mass as possible, enough torque to spin that mass up quickly enough for it to be effective before your opponent can get to you and stop it, and as much speed as possible to store more energy in the weapon. A spinning weapon is all about stored power!

A: Technical question - Mark J. here: Gyroscopic forces can cause odd effects for robots with a spinning mass weapon. Problems come when turning changes the orientation of a line running thru the weapon axle. With a horizontal spinning weapon, the axle points up and down, so turning the robot does not change the direction that line points -- no problem, as long as the weapon is well balanced.

With a vertical spinning weapon, turning the robot does change the orientation of the weapon axle. This results in forces that resist the turning motion and lift one side of the robot. The magnitude of the force is dependent on the speed of rotation, the mass of the weapon, and the radius of gyration. You can get full details on calculating gyroscopic forces at: www.freestudy.co.uk/dynamics/gyroscope.pdf

A: Different builders have different ideas about what's 'best'. It depends on your driving style and your skill as a builder. I see lots of different weapon types winning, but I notice robots with spinning drum weapons doing very well in the 12 to 60 pound classes.