Team 665 Fan/s and More

[Greg Peshek]

We're having fun with fans and conveyor belts, all in the same robot. Right now we've tested a 22", 3DA airplane prop. It puts out about 28 mph windspeed, and when we tried putting it on a rolling chassis alone that weighed about 90 lbs.. it started pushing it at about 3/4 speed and accelerated to a very nice speed at the end of 16ft. Here's a video of our fan testing: http://www.youtube.com/watch?v=Sw7jghN2rSU&feature=channel_page

Picture of our fan casing (also known as the giant cake dish):
http://farm4.static.flickr.com/3493/3242616363_6914769122.jpg

Picture of the fan casing being held up to the robot (this picture has all the conveyor stuff off)
http://farm4.static.flickr.com/3366/3242617185_7ac3602853.jpg

We tried the fan at a 1:6 ratio, we tripped the breaker before we could even get to a decent speed. We're looking at 1:1 as a very good ratio right now. We're looking at the possibility of two non-rotating fans for directional control. 2 weeks is plenty of time....

-Greg

[phr34kR]

Nice set up i like it, it seems like it has the potential to aid you in acceleration and handling good luck!

[Woodworker88]

Nice work. Out of curiosity, how much does a prop like that cost?

[EricH]

Nice work. Out of curiosity, how much does a prop like that cost?
If it's an R/C prop, it could be really cheap or really expensive. It depends on the type, whether it comes balanced or not, and where it's made. It varies. For example, my Aero team buys a particular brand of props. They're made in Australia, and shipping is kind of...well...not cheap. Then the props are carbon fiber. More not cheap. We do the balancing, so that saves a little money. But for your plastic, cheap type, it's not going to be nearly as expensive.

[Greg Peshek]

Nice work. Out of curiosity, how much does a prop like that cost?

We're using a XOAR wood laminated prop. Nothing particularly special, somewhere on the order of $30 per prop. While I'm not quite sure if they're balanced or not, it's certainly been quite smooth in all of our test runs.

-Greg

[EricH]

We're using a XOAR wood laminated prop. Nothing particularly special, somewhere on the order of $30 per prop. While I'm not quite sure if they're balanced or not, it's certainly been quite smooth in all of our test runs.

-GregFor better performance, try balancing it. If you don't have a prop balancer, all you need to make one is a couple of magnets, a steel rod, and something to stabilize the prop along the rod. Oh, and a freshman with fine sandpaper...

Find a place that is non-drafty. Set up the magnets just over a rod-length apart, and put the prop on the rod. "Hang" rod between the magnets, and give the prop a light flick. If it just turns, turns, turns, stops, it's balanced. If it stops, then reverses, then it probably isn't. lightly--very lightly sand down the heavy side for a few strokes. Repeat until the prop doesn't reverse.

Now, the motor doesn't have to fight against extra weight half the time. Overall faster prop...

[Woodworker88]

We're using a XOAR wood laminated prop. Nothing particularly special, somewhere on the order of $30 per prop.
-Greg

Wow that's significantly cheaper than I would have expected. I can't wait to see how it turns out when it's all done.

[Akash Rastogi]

Now that's a propulsion system :cool:

Very nice. Good luck and may the propulsive force be with you.

[Sean Raia]

Not JUST propulsion but possibly deflecting balls. :yikes:

[martin417]

I have a suggestion for measuring prop speed. The device you are using in the video looks for changes in light intensity. The room you are in has florescent lights. Florescent lights flicker 120 times per second. This flickering is picked up by the photo-tach and you will not get a good reading, if you get one at all. (try aiming the tach at the overhead lights, you should get a reading of around 3,600 RPM if you are set for a 2-bladed prop). You should make your readings outside, or turn off the overheads and use an incandescent light source like a flashlight, aimed from the far side of the prop (so the beam passes through the prop) at the tach.

On another note, you probably already know this, but static thrust (and the speeds you will be reaching in the game will be so low that static thrust is all you care about) is not affected by prop pitch, but power consumption is. So the trick is to swing the largest diameter, lowest pitch prop you can get. That will maximize the thrust you get.

For quick analysis, google thrusthp. Tis is a free program that will calculate power consumption, thrust and top speed based on prop diameter, pitch, and rotational speed.

Good luck at your regional!

[DarkFlame145]

idk why, but something about a propeller on a robot scares me =P But I like the idea

[sportzkrazzy]

One thing comes to mind when i c this. Air boat lol.
Have u guys thought about counter rotating props?

http://www.freepatentsonline.com/6821169-0-large.jpg

[gorrilla]

wow, i cant wait to see you guys at the regional:ahh:

[zrop]

We tried the fan at a 1:6 ratio, we tripped the breaker before we could even get to a decent speed. We're looking at 1:1 as a very good ratio right now. We're looking at the possibility of two non-rotating fans for directional control. 2 weeks is plenty of time....

-Greg

I'd suggest trying a 1:2 ratio. Because at peak power output you're gonna get about 2.6k rpm out of the CIM, and if you double that, for that kind of prop, you should be at a very good thrust. Also try to minimize your pitch, otherwise you'll be stalling the whole time. XD

[JesseK]

It puts out about 28 mph windspeed,

At some point you have to start wondering what all the backwash from your props will do to the other robots on the field, including your own partners. We'll find out in Orlando I suppose :rolleyes:

[Greg Peshek]

I have a suggestion for measuring prop speed. The device you are using in the video looks for changes in light intensity. The room you are in has florescent lights. Florescent lights flicker 120 times per second. This flickering is picked up by the photo-tach and you will not get a good reading, if you get one at all. (try aiming the tach at the overhead lights, you should get a reading of around 3,600 RPM if you are set for a 2-bladed prop). You should make your readings outside, or turn off the overheads and use an incandescent light source like a flashlight, aimed from the far side of the prop (so the beam passes through the prop) at the tach.

Good luck at your regional!

The device in the video is actually a wind speed meter. It has a little (muffin-like) fan in the center that spins and then it turns that into a wind speed. So we're not actually clocking the RPMs coming out of it, just the physical speed of the wind Thanks for the advice in the rest of the post, we'll make sure to look into it.

[Tom Line]

Back of the napkin tells me that if you're putting out a uniform 28 mph across a 32 inch fan, you'll get around 340 lbf out of it. Anyone want to check my numbers? I guessed on some of my constants - been a while since I cared about air density :rolleyes:

If my numbers are right, conservatively you may get 150 pounds of push out of it. Not too shabby!

[JamesByrne]

I do not have a link handy to the rules right now, but i thought somewhere in the rules you could not use a fan of any type to deflect balls that people are shooting at your trailer. So are these aimed sideways or are they aimed front and back?

Like i said i will come up with a quote when i get a miniuite to do so.

[Tristan Lall]

Back of the napkin tells me that if you're putting out a uniform 28 mph across a 32 inch fan, you'll get around 340 lbf out of it. Anyone want to check my numbers? I guessed on some of my constants - been a while since I cared about air density :rolleyes:

If my numbers are right, conservatively you may get 150 pounds of push out of it. Not too shabby!How much power would be consumed, in order to generate that? I don't think that airspeed is anywhere close to uniform over the propeller disc.

Anecdotally, I've dealt with a propeller-based system: a 100 lb model aircraft with a pair of 1 250 W (mechanical output) motors and Ø20 in two-bladed propellers with 12 in pitch. In long-duration maximum-power static thrust tests on the ground (there was resistance at the wheels, but not much), it had between 30 lb and 40 lb of thrust. The blades were spinning at over 6 000 rev/min, and the motors were drawing over 40 A each at 40 V. (The aircraft had a conservative flight duration of around 8 minutes at maximum power. When cruising, it needed far less power than that to sustain estimated airspeeds of 100 km/h.)

Incidentally, it's the experience on that project (everyone stayed safe), that fuels my healthy distaste for propeller systems in crowded spaces.

[Bongle]

I do not have a link handy to the rules right now, but i thought somewhere in the rules you could not use a fan of any type to deflect balls that people are shooting at your trailer. So are these aimed sideways or are they aimed front and back?

You can't use fan systems to de-score balls that are already in the trailer, but you are free to intercept incoming balls if you should so please.

I am looking forward to seeing these fan-bots in action.

[Tom Line]

How much power would be consumed, in order to generate that? I don't think that airspeed is anywhere close to uniform over the propeller disc.

Anecdotally, I've dealt with a propeller-based system: a 100 lb model aircraft with a pair of 1 250 W (mechanical output) motors and Ø20 in two-bladed propellers with 12 in pitch. In long-duration maximum-power static thrust tests on the ground (there was resistance at the wheels, but not much), it had between 30 lb and 40 lb of thrust. The blades were spinning at over 6 000 rev/min, and the motors were drawing over 40 A each at 40 V. (The aircraft had a conservative flight duration of around 8 minutes at maximum power. When cruising, it needed far less power than that to sustain estimated airspeeds of 100 km/h.)

Incidentally, it's the experience on that project (everyone stayed safe), that fuels my healthy distaste for propeller systems in crowded spaces.

Like I said - 320 pounds if uniform. Assume it's not, so I halved the value to 150. Even if you want to cut it 80%, they're going to get similar drive from that fan as they do from the wheels, so they'll accelerate twice as fast as a non-ducted bot.

They better not have the air-intake on the bottom of the robot though.....

[wendymom]

I'm a big fan of 665...tee hee. Actually I think this is going to be really fun to watch

[zrop]

I like how no one is really that concerned about safety here when the guy has his hand REALLY close to the prop, yet when our team posts anything about prop testing, we are continually nominated for the Darwin award XDD

[sdcantrell56]

his hand is close to the blade but not in the plane of rotation like you guys were. Also there protection around the fan is far superior to the one you have currently proposed. I think people are becoming tired of your blatant disregard for safety. It's ok if you want to hurt yourself, but by not shielding the fans properly you will be putting hundreds of people at risk.

[spc295]

I think the safety officails will have fun with this team:rolleyes:

[Ryan Caldwell]

I like it, great concept, can't wait to see it in action!

[Woody1458]

I like how no one is really that concerned about safety here when the guy has his hand REALLY close to the prop, yet when our team posts anything about prop testing, we are continually nominated for the Darwin award XDD

It has a lot to do with the way your post came off. It has been talked about intesely else where. You should stop considering yourself a "victim" but rather extremely assisted. We all want you to pass inspection and these are the steps necessary to make that happen.

[Greg Peshek]

Like I said - 320 pounds if uniform. Assume it's not, so I halved the value to 150. Even if you want to cut it 80%, they're going to get similar drive from that fan as they do from the wheels, so they'll accelerate twice as fast as a non-ducted bot.

They better not have the air-intake on the bottom of the robot though.....

During the video I posted, you can see our mentor measuring the wind speed at different locations around the blade. It is roughly 26 mph at the far tips of the blade and 28 mph toward the middle of the blades. Don't worry, the air-intake on the bottom was our original plan until we pulled up our test flooring on the third day. Then we saw the update and that was categorically disqualified.

At some point you have to start wondering what all the backwash from your props will do to the other robots on the field, including your own partners. We'll find out in Orlando I suppose

There's a big volume of air being pushed out, if you watch the video closely, you can see the papers blowing 20' away from the prop, but we are fairly sure by the time the backwash diffuses after being pushed out the fan there shouldn't be any giant surfaces for the air to catch on with the other robots. Let's just hope no one puts sails on their robot. :)

For anyone interested, we measured the amp draw at different power outputs.. it's about 20 amps on startup, 8-10 amps on 50%-75% speed, and 4 amps at 100% throttle.

As far as the safety aspect, the metal for the cowling is only 1/32", so it was fairly easy to bend and make into a circle. Although we are confident that if the blade is to come off and shatter, the cowling would stop or drastically reduce the velocity of any dangerous debris. Since there really isn't any scale, the mesh we have protects against putting fingers in. One girl on our team can fit her pinkie finger in the mesh, but even then it wouldn't be long enough to get to the blade. So we figure a 10 year old won't be able to hurt themselves even if they tried.

Thanks for all of the comments and input guys.

-Greg

[Akash Rastogi]

I like how no one is really that concerned about safety here when the guy has his hand REALLY close to the prop, yet when our team posts anything about prop testing, we are continually nominated for the Darwin award XDD

Keep it in your own thread please.

[JamesByrne]

You can't use fan systems to de-score balls that are already in the trailer, but you are free to intercept incoming balls if you should so please.

I am looking forward to seeing these fan-bots in action.

What rule states this? Sorry i'm just trying to get clarification.

[EricH]

What rule states this? Sorry i'm just trying to get clarification.
Q&A in http://forums.usfirst.org/showthread.php?t=10943

[JamesByrne]

Thank you very much it settled an argument. I appreciate your fast response.

[writchie]

Back of the napkin tells me that if you're putting out a uniform 28 mph across a 32 inch fan, you'll get around 340 lbf out of it. Anyone want to check my numbers? I guessed on some of my constants - been a while since I cared about air density :rolleyes:

If my numbers are right, conservatively you may get 150 pounds of push out of it. Not too shabby!

You're off by about one order of magnitude.

Spinning a 22 inch prop at 4000 RPM is going to get you a little over 11 lbf of static thrust.

Forget about windspeed. Your goal is not to move air. It doesn't matter how fast you move air. The static thrust from your propeller (i.e. what moves the robot) is a function of prop diameter, number of blades, and RPM. The propeller pitch will determine how much torque (and therefor mechanical power) is required. Nothing else matters much.

The most mechanical power you can get out of a single CIM is about 300W. While a pair looks like it could get you 600W, you will have serious voltage drops so 500W is a more realistic maximum mechanical power for 1 to 2 second intervals with 60 amps on each motor. If you want to operate sustained, you have to limit your input current to not much more than 40 amps each. At the max power point 60% of your input power is going to go into motor heat. Your continuous mechanical power is unlikely be more than 440W for the pair. You'll loose at least another 5% in your gear train.

You could theoretically reach about 21 lbs of thrust with a 22 inch prop spinning at 5500 RPM with a pair of CIMS geared about 1:2. But to do this you will need a prop pitch under 2. A normal prop pitch of 4 is just going to go click - click - click because it's torque load at 5500 RPM requires more than 800W and you don't have that kind of power. You'll top out at 4000 RPM and about 11 lpf static thrust.

BTW We are using a pair of FP motors each driving a 12 inch 3 blade adjustable prop. Our CIM's are dedicated to other uses and call me old fashioned but I just don't like the idea of gearing up a motor. (Although it appears unavoidable if you want to use CIM's).

The "best practice" way to determine your static thrust is with an engine test stand setup that measures the actual force exerted at the prop shaft. R/C airplane folks do it all the time. Google is your friend.

The above numbers are for open air at sea level. The safety shroud will normally have a bit of a positive effect and your safety cage will have a negative effect. Any negative pressure created by inadequate air flow behind the prop will have a (possibly severe) negative effect as well. And if you plan to compete at the Colorado regional, you'll loose about 20% to Denver's density altitude ;)

[NOV8R]

I fly large (25 lb) radio control airplanes with 52 cc gas engine that swing 20 x 10 props at 7000 RPM static. The engines idle at just below 2000RPM. Your not going to get much thrust using one of these props at the 2600 RPM that a CIM turns.

[zrop]

You biggest problem is going to be finding a large prop like that, that has a pitch of 2 or less. At this point, you should either try to custom order one, or get a nice piece of wood.. and carve your own =)

[travis]

I am not an aeronautical engineer, but I do sit next to one. I also attended the AE interweb school of hard knocks when I decided that work well or not, propellers are going on the 2091 machine. Here’s what I found.

Your standard RC airplane propeller is designed to run at high RPMs AND at high forward velocity. Of course they work stalled, lest a plane wouldn’t take off, but their peak efficiencies are when they are moving. I don’t like the idea of gearing up a motor either, and the 5krpm type props were too big for what I was going for. They make propellers designed to run on motors (not engines) with typically higher, low-RPM torque requirements and lower total RPM sweet spots. I was poking around looking at those when I found the so called “slow fly” propellers. These are designed to fly at low RPM, and in stalled conditions. APC propellers makes FRP slow fly propellers in left and right handedness -- so you can have a counter rotating arrangement -- in lots of pitch and diameter combinations, surprisingly (to me) cheaply. The ones I settled on were 12x3.8 through dragan fly innovations. They are marketed for dual rotor RC helicopters in the 300W motor range, which looked perfect for cim motors. The slow fly propellers maximum RPM is about 65,000/D[=]in, so for 12” props it is ~ 5500, which I should never see in a loaded cim. This was one more piece of evidence that this was the right propeller/motor combination. A couple of suspect online propeller calculators indicated that each propeller would be in the neighborhood of half the motive force of what the wheels could punch out, which is testing enough for me. I’m bring’n the Heavy Metal (well, FRP) Noise to New Orleans (world capital of airboats), whether they work or not; 2024-T3, solid rivet construction enclosures by the way, before anyone jumps all over me.

Any real AEs see anything wrong with that?

Travis

[writchie]

I don't think you will be anywhere near 5 lbf per CIM with a slow fly prop, although I'd love to be wrong about this.

For reference, A Draganflyer X6 Helicopter has 3 sets of 16inch/15inch contra-rotating blades, i.e. 6 blades operating at 2000 RPM at hover with a maximum gross weight of 3.3 lbs (and max power for its 6 motors of 450 watts total. So if you can beat this by 50% with a single smaller prop they will probably have a job waiting for you ;)

The mechanical power you can get from the CIM operating from 40 - 50 amps is about 125 - 185W. But you can get that only with a torque load of 100 - 125 oz-in. You need a prop that will give you that torque load at 3800 to 3300 RPM respectively.

Slow fly props do have higher loads than normal props but I suspect that a 12 x 3.8 is still pretty far from the sweet spot on a CIM. A normal clark airfoil type prop is only going to give you about 1.2 - 1.5 lbf at 5000 RPM.

BTW, a traditional 15 x 3.8 prop at 2000 RPM would produce about .61 lbf. So 6 of them is within 10% of the max gross of the X6. Drop the prop to 12 inches and you've got a quarter pounder.

I've seen test data showing an APC 10x3.8 slow fly turning at 6850 RPM producing 1.54 lbf static thrust which is pretty good agreement with 1.69 lbf for a normal prop. Note that this is already above the max recommended speed. Slow fly props are not as sturdy as conventional props. You won't get the CIM anywhere near 6850. I think you'll see 1 - 1.5 lbf.

I think the major difference in the slow fly is that you will have better performance at slower speed but nowhere near the improvement of dropping the pitch under 2.

Once the robot gets moving, the effective pitch drops and the slow fly should increase faster to its max thrust point. The normal prop will never get to that point. The lower pitch prop is starting out well ahead of the slow fly and will also improve.

It should be interesting to see what end's up working out the best and what can survive the abnormal precession forces resulting from robot collisions. This is unexplored territory.

[martin417]

Based on some preliminary, very unscientific testing* with bare props, directly mounted to Fisher Price motors, in pusher configuration, we were getting ~2.0 - 2.5 Lb.s of thrust from APC slow fly 10 x 3.75 props. They were running at about 9,000 RPM. If you look at this paper (http://aerade.cranfield.ac.uk/ara/1958/naca-tn-4126.pdf) (yes it is old, but the science is still good) you will see that it is possible to gain about 50% to 75% in thrust, as well as improved efficiency by building a shroud. Since you have to protect the props in something for safety anyway, why not make it an asset? Here is a picture of the fiberglass shroud we made for the prop:

http://picasaweb.google.com/mwilson417/First#5299287211437329698

Link: (http://picasaweb.google.com/mwilson417/First#5299287211437329698)

On another note, due to the internal resistance of the battery and other components, if we ran two of these at the same time, the voltage dropped enough that the RPMs fell off to about 8,000, and the thrust went down to less than 2 lb. per prop.

We are building these modules, but still don't know whether or not we will add them to the finished bot. We still have to do testing to see whether the benefits outweigh the hassle and if they fit within the weight budget.

*Unscientific test: stand on digital scale holding motor/prop in hand and record displayed weight. Point the prop straight down and engage motor (use caution here, very dangerous, do not attempt this at home! these are trained professionals very stupid people!). record displayed weight again. Calculate difference, This is your thrust.

[travis]

So, if a 12x3.8sf propeller is not the best candidate, what is a better one? Let’s make the design criteria D<=12” and a cim with 1:1 transmission. I had a hard time finding pitches below 3 in the 12” class, especially ones that didn’t run at the more like 15Krpm nitro speeds. Would an APC 12x10E thin electric be more my speed maybe? If there isn't a good option in the 12" range, I could maybe squeze into a size 14.

I have a working drive train, so the propellers are just gravy at this point. Even a modest 1/4lb thrust is justification enough for installation in my opinion.

Thanks,
Travis

[martin417]

So, if a 12x3.8sf propeller is not the best candidate, what is a better one? Let’s make the design criteria D<=12” and a cim with 1:1 transmission. I had a hard time finding pitches below 3 in the 12” class, especially ones that didn’t run at the more like 15Krpm nitro speeds. Would an APC 12x10E thin electric be more my speed maybe? If there isn't a good option in the 12" range, I could maybe squeze into a size 14.

I have a working drive train, so the propellers are just gravy at this point. Even a modest 1/4lb thrust is justification enough for installation in my opinion.

Thanks,
Travis

Theoretically, a CIM should be able to swing that 12 x 3.8 prop at 8500 rpm, giving you 4 1/2 lb. thrust, and only using .35 HP. Sounds like a good candidate to me.

[writchie]

Theoretically, a CIM should be able to swing that 12 x 3.8 prop at 8500 rpm, giving you 4 1/2 lb. thrust, and only using .35 HP. Sounds like a good candidate to me.

The no load RPM spec on the CIM is 5310 @ 12V. Trying to remain 1:1 is the basic problem. I don't see how you can do it without a much larger dia. At a torque load and RPM that will give you the required shaft horsepower for this prop at 8500 RPM you would need to gear it up about 1:2.75. But those numbers are really for a normal prop, not a slow fly. SF props at this size are not safe at this RPM even for airplanes. When you add the likelihood of shock induced tip flutter from collisions it's a bit scary.

We are using direct drive FP motors and our thruster is still optional. We were originally planning to machine a custom prop but are now planning on off-the-shelf adjustable pitch props with a custom machined 3 blade hub to get us a 1 - 2 pitch in a pusher configuration.

As another poster pointed out, careful attention to the shroud will help.

I suspect we may see both geared up CIM's and Direct Drive FP's with total thrusts approaching 10 lbf (at the start of the match).

[writchie]

If you look at this paper (http://aerade.cranfield.ac.uk/ara/1958/naca-tn-4126.pdf) (yes it is old, but the science is still good) you will see that it is possible to gain about 50% to 75% in thrust, as well as improved efficiency by building a shroud. Since you have to protect the props in something for safety anyway, why not make it an asset? Here is a picture of the fiberglass shroud we made for the prop:

http://picasaweb.google.com/mwilson417/First#5299287211437329698

Link: (http://picasaweb.google.com/mwilson417/First#5299287211437329698)



Good stuff. Careful what you call old ;).

Be aware that 9,000 RPM is about 138% of the manufacturers safety limit for a 10 inch slow fly prop.

The battery issues (and wiring) are very important. While the internal resistance of the battery is spec'd at 10 milliohms it behaves more like 5 to 10 times that at higher currents and low frequencies. It's best to design around 11 volts at the motors to account for all of the drops.

[Urban Hawk]

i got the rule that states you can't de-score game pieces already in the trailer.

<G25.1> De-scoring GAME PIECES – Once a GAME PIECE has been SCORED, it may not be intentionally de-scored (e.g. removed from the TRAILER). De-scoring a GAME PIECE will cause a PENALTY to be assigned. At the end of the match, any intentionally de-scored GAME PIECES will be considered SCORED as originally placed. GAME PIECES that are knocked free from tenuous placements as a result of normal game interactions (e.g. a GAME PIECE on top of a pile of MOON ROCKS that completely fill a TRAILER falls off when the TRAILER is bumped) will not be penalized.

also i think you should angle the fan slightly so it provides some extra downward force and it would probably help to deflect the balls better then putting the fan horizontal because the balls would practically be in the trailer by the time it hits the wind from the horizontal fan while the angled fan should hit the ball while it is still further up and more of a chance to stop it from going in since the balls are mainly going to be lobbed in.

[EricH]

also i think you should angle the fan slightly so it provides some extra downward force and it would probably help to deflect the balls better then putting the fan horizontal because the balls would practically be in the trailer by the time it hits the wind from the horizontal fan while the angled fan should hit the ball while it is still further up and more of a chance to stop it from going in since the balls are mainly going to be lobbed in.
Sorry, but Q&A got to that one pretty quickly when it was brought up. http://forums.usfirst.org/showthread.php?t=11025

Yeah, it's intended to deflect incoming moon rocks. But I can almost guarantee you that at least one inspector per event will not allow the up-angle. (Most likely the lead inspector, but could be another one.) It's going to be really hard to build a deflection system that works using air jets, because it has to be horizontal. That said, anyone who does a fully legal one deserves at least a Xerox Creativity Award.

[Greg Peshek]

Sorry, but Q&A got to that one pretty quickly when it was brought up. http://forums.usfirst.org/showthread.php?t=11025

Yeah, it's intended to deflect incoming moon rocks. But I can almost guarantee you that at least one inspector per event will not allow the up-angle. (Most likely the lead inspector, but could be another one.) It's going to be really hard to build a deflection system that works using air jets, because it has to be horizontal. That said, anyone who does a fully legal one deserves at least a Xerox Creativity Award.

Yeah and the thing is, we tried dropping the ball right in front of the full spinning prop that you can see in the video and it moved maybe an inch? Even quadrupling the air output wouldn't do enough to deflect it away. Now if you could magically get the camera to track incoming balls and have a redirect able air funnel from the prop to shoot at the ball.. you could do it. But, you might have a lot of fun getting that one to work.

-Greg

[jwkelly]

Back of the napkin tells me that if you're putting out a uniform 28 mph across a 32 inch fan, you'll get around 340 lbf out of it. Anyone want to check my numbers? I guessed on some of my constants - been a while since I cared about air density :rolleyes:

If my numbers are right, conservatively you may get 150 pounds of push out of it. Not too shabby!

I'd have to say your estimate is quite generous. We have run a fan test on a 28 inch prop with it directly connected to a CIM motor (1:1). Xoar 28x10 prop gives us approximately 8lb static thrust, whilst a 28x12 gives approx 7lb of static thrust.

[Urban Hawk]

sorry, but Q&A got to that one pretty quickly when it was brought up. http://forums.usfirst.org/showthread.php?t=11025

Yeah, it's intended to deflect incoming moon rocks. But I can almost guarantee you that at least one inspector per event will not allow the up-angle. (Most likely the lead inspector, but could be another one.) It's going to be really hard to build a deflection system that works using air jets, because it has to be horizontal. That said, anyone who does a fully legal one deserves at least a Xerox Creativity Award.

well i looked at the thread you gave me but id rather have to dissagree with it. There are 2 main arguments in there against it. The first being the vaccume will damage the floor and the second being it goes against <R06>.

For the first of the 2 I'd have to say that the main argument there is merely about suction such as a vacuum cleaner would provide. How ever if you have it so that the fan draws the air from a large area around it then it shouldn't create a vacuum and should dispel that particular argument.

As for the second with it breaking rule <R06> from what i see of the rule it isn't breaking it.
<R06> ROBOTs must use ROVER WHEELS (as supplied in the 2009 Kit Of Parts and/or their equivalent as provided by the supplying vendor) to provide traction between the ROBOT and the ARENA. Any number of ROVER WHEELS may be used. The ROVER WHEELS must be used in a “normal” orientation (i.e. with the tread of the wheel in contact with the ground, with the axis of rotation parallel to the ground and penetrating the wheel hub). No other forms of traction devices (wheels, tracks, legs, or other devices intended to provide traction) are permitted. The surface tread of the ROVER WHEELS may not be modified except through normal wear-and-tear. Specifically, the addition of cleats, studs, carved treads, alterations to the wheel profile, high-traction surface treatments, adhesive coatings, abrasive materials, and/or other attachments are prohibited. The intent of this rule is that the ROVER WHEELS be used in as close to their “out of the box” condition

The fan is not a traction device because it doesn't provide traction. It merely causes an increase of the force caused by frame of the robot (and the components of the robot) in pushing/pulling downward on the wheels. The robot already naturally does this through gravity but the fan just increases it further and so increases the wheels natural friction without any form of modification to the wheels.

[EricH]

As for the second with it breaking rule <R06> from what i see of the rule it isn't breaking it.


The fan is not a traction device because it doesn't provide traction. It merely causes an increase of the force caused by frame of the robot (and the components of the robot) in pushing/pulling downward on the wheels. The robot already naturally does this through gravity but the fan just increases it further and so increases the wheels natural friction without any form of modification to the wheels.Let's not start that debate again. It came up early in the season when someone first suggested fans for propulsion. I don't remember the exact thread, but a quick search should turn it up.

I will simply say that the apparent intent of <R06> is that you cannot increase your traction by any means other than a heavier robot, and getting a heavier robot by adding a fan to increase your normal force is illegal according to the Q&A. I just wish that they'd make that clear as glass and end all these debates.

Incidentally, your physics is wrong. µ (the coefficient of friction) never changes, so you can't say that friction increases. However, Ff (the frictional force) = µ*N, where N is the normal force. Because µ is constant, if you increase N, you increase Ff. Friction never increases; frictional force does.

[robert2.0]

tere must be a hopper but where>?:confused:

[Urban Hawk]

I think i know which thread you were talking about. The one i found was:Team 2526 - Propeller Propulsion Prototype. As well as that post on the Q&A. i guess that you are right about not being able to use it to provide a downward force.

[Greg Peshek]

tere must be a hopper but where>?:confused:

You can see in this more recent picture that the balls are stored in the conveyor, we run into people, and dump them out into the trailer. We can store 8 balls and dump them all in 1-2 seconds. We can also eat people's bumpers, it's fun. The design is very similar to the design that 1712 posted on here. We just plan on having two separate towers, so instead of running into the problems where the balls get all jumbled and stuck in the conveyors, there is a constant column of them.

http://farm4.static.flickr.com/3327/3260971993_9c2ef4c946.jpg

There is now dividing lexan to keep the balls separate that isn't shown in that picture. The fan/s is/are going right behind that tower, over the electronics.

We've tried multiple pulley setups/ one versus two CIMs and at this blade size and weight the pulleys actually start becoming inefficient from the vibrations and loss in the pulleys. We have found a way to increase thrust (not windspeed) by 20-30% on only one CIM with no gear ratio. I'll post more pictures and videos on the whole setup near the end of build season.

The idea is to be able to get all 8 balls in, turn on the fan to zip across the field, violently run into someone (with the aid of the fan), dump, repeat. We're hoping the space lost in the manipulator design to the fan will be made up for by the advantage of the fan. If we can use the fan to catch people/ push easier.. we believe we'll have an advantage over hopper dumpers/ power dumpers that may have a harder time of pinning.

We'll see..it'll be an interesting year for us to say the least.

-Greg

[EricH]

I think i know which thread you were talking about. The one i found was:Team 2526 - Propeller Propulsion Prototype. As well as that post on the Q&A. i guess that you are right about not being able to use it to provide a downward force.Actually, I was referring to one from even earlier, something about propulsion without using wheels. There was a big debate until the above referenced Q&A came out.

[boomergeek]

We were also experimenting with props direct drive off the FP motors. We tried APC 11x3 spinning at 7800-8000 RPM and it produced under 2 lbs in our static thrust test. When we tried the 11x6, it produced about the same thrust at a lower RPM but the same power. ThrustHp indicated that the 11x3 should produce a 50% better thrust than the 11x6. We've found other analysis tools (http://www.mvvs.nl/prop-power-calculator.xls) that indicate that as RPMs go up that the hp to thrust ratio goes down and tends to cancel the finer pitch to thrust ratio improvement. This "prop-power-calculator" seems to match our test results closer than the ThrustHp program. I'm curious if anyone have found that the ThrustHp is seemingly too optimistic for pitches under 10 degrees.

We are still searching for the "optimum" 11" prop: either off the shelf or fabricating ourselves (Ooooh, the mental energy we expend to eek out extra pounds of thrust...).
It's probably too late to study to become an airfoil designer- I'd settle for trying to be an airfoil hack. ;)

The no load RPM spec on the CIM is 5310 @ 12V. Trying to remain 1:1 is the basic problem. I don't see how you can do it without a much larger dia. At a torque load and RPM that will give you the required shaft horsepower for this prop at 8500 RPM you would need to gear it up about 1:2.75. But those numbers are really for a normal prop, not a slow fly. SF props at this size are not safe at this RPM even for airplanes. When you add the likelihood of shock induced tip flutter from collisions it's a bit scary.

We are using direct drive FP motors and our thruster is still optional. We were originally planning to machine a custom prop but are now planning on off-the-shelf adjustable pitch props with a custom machined 3 blade hub to get us a 1 - 2 pitch in a pusher configuration.

As another poster pointed out, careful attention to the shroud will help.

I suspect we may see both geared up CIM's and Direct Drive FP's with total thrusts approaching 10 lbf (at the start of the match).

[martin417]

We did a lot of experimentation and comparison with ThrustHP. While the program was a little optimistic, it was not found to be overly so. Regardless of program or testing, static thrust is completely independent of pitch. Use the lowest pitch prop you can get for a given diameter (power is affected, so lower pitches turn faster, giving more thrust) We used two APC 10 x 3.75 Slow-Flyer props, mounted in a custom fiberglass shroud and direct driven with Fisher Price motors. When measured with a fish scale, the setup produced 6-8 LB.s of thrust on average. The fans were an optional "booster", activated by the joystick trigger (only if the joystick output was greater than .4) They were some help in a pushing match, but the real boost came when chasing a robot for the score. When the driver hit props, you could see the bot accelerate down the field. We called it "turbo boost" mode.

[boomergeek]

Thanks for your analysis. I had the idea from reading your previous posting (http://www.chiefdelphi.com/forums/showpost.php?p=814537&postcount=38) that if you ran both fans at the same time that you were getting less than 2 pounds of thrust each due to battery loading. Your previous analysis seems to correspond to the numbers we were getting with a pair of 11x3s.

If you don't mind me asking, how did you achieve jump from under 4 pounds to the 6 to 8 pound range?

We will likely try some experiments with a 10x3.8 SF and a more optimal shroud. We were also experimenting with "javaprop" to understand what the optimal airfoil shape would be.

[sdcantrell56]

I don't know all the details but I believe our shrouds increase our thrust by about 150%. I know that we greatly out accelerated everyone and could push multiple robots at once at will.

[martin417]

Thanks for your analysis. I had the idea from reading your previous posting (http://www.chiefdelphi.com/forums/showpost.php?p=814537&postcount=38) that if you ran both fans at the same time that you were getting less than 2 pounds of thrust each due to battery loading. Your previous analysis seems to correspond to the numbers we were getting with a pair of 11x3s.

If you don't mind me asking, how did you achieve jump from under 4 pounds to the 6 to 8 pound range?

We will likely try some experiments with a 10x3.8 SF and a more optimal shroud. We were also experimenting with "javaprop" to understand what the optimal airfoil shape would be.

In the previous post, I mentioned the "very unscientific (and dangerous) testing" and the "bare props" In the more recent post, I was talking about the same props, mounted in a custom fiberglass shroud, designed upon a paper written by NACA (see this thread for details) (http://www.chiefdelphi.com/forums/showthread.php?t=72863&highlight=naca) and testing on the bot with accurate fish scales. The shroud can theoretically give a 50% to 75% boost in thrust. So, in answer to your question, The first numbers may have been off by good bit due to the resolution of the bathroom scale, the shroud may have done the trick, or some other unforeseen factor. At any rate, I am confidant in the latest data due to much more rigorous, scientific methods of test.

[boomergeek]

Thanks for clarifying. . .I'll also boost our Ae to be closer to 1.6 of Ad. The first test shroud (cut from a 5 gallon pickle bucket and with a garden hose for the radius of the lip) had Ae about 1.2 of Ad.

We are working to improve the rigorousness of our testing methods.

[martin417]

Thanks for clarifying. . .I'll also boost our Ae to be closer to 1.6 of Ad. The first test shroud (cut from a 5 gallon pickle bucket and with a garden hose for the radius of the lip) had Ae about 1.2 of Ad.

We are working to improve the rigorousness of our testing methods.

According to the NACA paper, the parameter that had the most effect on thrust was inlet radius. The radius should be at least 6% of prop diameter, more is better (ours is 15%, or 1.5" radius on a 10"prop) Ae and Ai had some effect, as did shroud length (we had to cut about 1.5" off our shroud in order to get it to fit on the bot, and noticed no reduction in thrust). These effects were small in comparison to inlet radius. For more info, the paper can be found here (http://aerade.cranfield.ac.uk/ara/1958/naca-tn-4126.pdf).

[boomergeek]

We did try to understand the paper (reading it several times) but we did not do enough fabrication and experimentation. From the discussion associated with Figure 5 and 6, we thought that the advantage of larger lip radius beyond .06 toward .15 would not provide a significant advantage. The paper did not indicate the effects of the size of the lip radius AFTER reaching 90 degrees from the general thrust direction: I.e., how large surface plane (i.e., what the paper describes as the simulated upper surface in figure 1)needs to be after the radius reaches 90 degrees to achieve the lip radius benefit. On our inital tests, we assumed we could extend the lip radius to 180 degrees(basically following the curve of the hose): I think that is probably our larger problem.

We are limited in how big we can make the "simulated upper surface" to about a 12" to 13" square for each shroud.

Thank you for your continued expert advice for trying to help us to a more optimal configuration.

The amount of team energy to work on upgrades for our second and last regional is not as strong as the enthusiasm back before the first regional.

We will definitely let you know how it turns out.

[martin417]

We did try to understand the paper

The paper isn't terribly clear, but we concentrated on a few graphs, mainly the T/Tu graph and T/inlet Radius. We made the inlet radius 1.5, giving us a 13" diameter inlet (we went big because we knew we would be obstructing it some with guards) we used a 1" straight section where the prop runs, and about 4" of 7° angled outlet. We had to cut off some of the outlet to get it fit, but as I said, it didn't seem to affect the thrust. The straight section was a little less than 10" so we spun up the prop and touched it with a large flat file to get tip clearance (you want the tip to be as close as possible to the shroud, be very careful here! use safety shields and go slowly!)

here are some pictures of the finished product:

Shroud (http://picasaweb.google.com/mwilson417/First#5299287211437329698)

Mounted to bot (http://picasaweb.google.com/mwilson417/First#5314157348509577666)

[boomergeek]

The shroud based on the NACA report worked great for us.
Special thank you to Martin417 of team 1771 for advising us with the finer points to get it right. We also added a twist to the whole fan concept by putting a pair of fans on a turret. By the end of our matches at the Connecticut Regional, our drivers and copilots were doing amazing coordinated maneuvers that had the announcing play-by-play guy very impressed. The vectored thrust from the fans was really a fun innovation: I just wish we had more time to practice with it.

http://i652.photobucket.com/albums/uu247/boomergeek/robotwithfans.jpg
http://i652.photobucket.com/albums/uu247/boomergeek/propsinshroud.jpg

If we could start over with this game, I would suggest, all that is needed is a turreted stack of properly shrouded fans on top of any four wheel drive system. Then simply push and pin the best robot on the other alliance sequentially in front of each of the three human alliance shooters.

I think an alliance of 3 turreted stack of fans bots would humiliate any alliance of offense-based robots. The offensive based bots might be able to score their first 7 balls, but they would be incapable stopping the turreted fan bots from quickly pushing them to the fan bots human shooters and then the fan bots could easily hold them there for the full match.

[martin417]

Kudos on a beautiful implementation and a neat "twist" on the fan idea! Are you going to the championships? If so, I will definitely come by and take a look in person.

[boomergeek]

Kudos on a beautiful implementation and a neat "twist" on the fan idea! Are you going to the championships? If so, I will definitely come by and take a look in person.

We would love to be making plans to go to the championships this year, but its not in the cards. Too many excellent robots: three heartbreak losses at the buzzer; (But our hearts mended quickly and we were all dancing in the stands and using giant letters to spell out cheers for the remaining teams.)

It was my first year involved and I really came away impressed with the generosity and talent of the students, mentors and volunteers.

I hope to be involved for many years to come.

Our team got to go to the championships last year and we'll start working on the goal of getting back there next year.

Thanks again and good luck at the championships.