All Terrain Vehicle

[aswincool]

Hi Guys i have read various discussions in this forum, am doing all terrain vehicle as my final year project.. i have read the thread "high torque in dc motor".. The comments helped me a lot in understanding the torque concept in flat, inclined and unstructured surfaces..

am going to use 60 rpm, 12 V gear head motor with torque of 38 kg cm

My question is if am designing a 4 wheeled robot with 4 pully wheels and belt, is it enough to use the above rated motor for the front two wheels?

 

[dknguyen]

Not if your robot weights 500kg...or if you use 1 meter diameter wheels...or if you use the belt/pulleys to change the decrease the gear reduction ratio so there isn't enough torque...or if you want to scale steep inclines...or if you use really soft grippy wheels... You should know by now to provide WAY more detail.

https://www.electro-tech-online.com/threads/motor-sizing-for-moving-robots.23264/#post155131

Remember to take into account the pulleys wheel diameters being able to change the gear ratio.

 

[aswincool]

Thanx dude..
I did the calculations u said and i hav calculated the torque per motor as 3.53Kg

for atv i hav considered tank model

In tank model more number of wheels are attached to the belt, If am using 2 21 cm wheels and 2 6.6 cm wheels on a single side and a belt linking all the above should i use same rating[rpm,torque] motor to drive all its wheels?

 

[aswincool]

Which one would be a better design for the ATV?

Which one would be a better design for the ATV?

 

[dknguyen]

The one on the left and middle are about the same except the middle can have superior climbing ability but can't be inverted (unless you make it a parallelogram) The one with the arms is mroe capable of climbing a larger step for the same profile and can flip itself if its upside down but is much more complicated to make. it needs very strong servo motors and a complex mechanism to drive teh arms on the track that can also rotate the arm independently at the same time You'll have a hardd enough time pulling off the design on the left. The moveable climbing track seems in a place different place than is usual though:

iRobot Ground Robots - 510 PackBot
**broken link removed**

 

[aswincool]

Such a vehicle is our source of inspiration but i still can't find a way to control two wheels[support arm wheel+ regular wheel] on a same shaft independently

 

[dknguyen]

For a track, what matters is how large the sprocket is that the motor uses to drive the track. A larger sprocket is like a larger wheel- more diameter means more speed but less force. So to move the same robot of the same weight at the same speed, a larger drive sprocket needs a motor geared for more torque and less RPM and a smaller drive sprocket needs a motor geared for more RPM but less torque. Both will get the same results- it's like the size of the drive sprocket is part of the gearbox. The size of the idler pulleys that simply support the belt don't affect the torque required. They just press the belt on the ground so it doesnt flex over bumps and keep the belt from rolling off the frame.

3.5kg of torque per wheel? Or do you mean 3.5kg of force per wheel? Because torque has units of force-distance, and the same robot needs the same force per wheel no matter what the diameter of wheel is, but as the diameter gets larger, you need more torque to produce the same force on the edge of the wheel (because the radius which is the lever arm is longer). You get more speed though with larger wheels (but less force) because a larger wheel spinning at the same RPM moves farther faster than a smaller wheel spinning at the same RPM. You also get better climbing ability.

But in a simple straight track (like your design on the left), climbing ability is affected by the diameter of the leading pulley/sprocket (just like a wheel). But in a shaped track (like your design in the middle) it is affected by the height of the leading ramp. This leads to an important advantage of tracks- in a shaped track you can make it look like a very large diameter wheel by using a very large ramp but dont need a motor that is geared down by a huge amount because you can use a very small drive sprocket to turn the belt (whereas for wheels you suddenly have a very large radius which is a very long lever arm that needs lots of torque). Moving track arms (like your design on the right) simulates an adjustable shaped track (with a lot more complexity)

All tracks benefit from not getting hung up or stuck as easily as wheels but have high friction turning.

Honestly though...the design on the right with the movign arms is out of the league of anyone who isn't a machinist or cant truly make their own mechanical parts. Even between the left and middle designs, you're probably stuck with what you can pull off. Moving track arms is heavy, bulky, expensive, and complicated and I wouldnt try it. Tracks have enough little parts as it is.

Notice that the second robot I linked to just has arms- there are NO TRACKS on the arms. It just helsp the robot lift itself up a step that is taller than its leading edge sprocket will allow.

 

[aswincool]

" Just center it vertically in the middle of the robot so it is symmetrical upside down and right side up." what does it mean? am unable to imagine..

BTW 3.5kg force and am using 21cm wheel and torque is 38kg which am using

I dont think it's of same motors as 1 ctrl is for wheel rotation and the other for support arm's angle movement..

I remember u said abt idealr wheels, what if i drive them them also by 20 rpm 4 kgcm torque motor.. would it help me drive the track?

 

[dknguyen]

" Just center it vertically in the middle of the robot so it is symmetrical upside down and right side up." what does it mean? am unable to imagine..

Please ignore that comment. I have deleted it. I thought about it more and take it back. Your design on the right takes extra work and doesnt work as well and is heavier compared to the I-Robot designs. The attached PDF is what I meant to describe though.

" I dont think it's of same motors as 1 ctrl is for wheel rotation and the other for support arm's angle movement..

When I said separate motors I mean separate motors to spin the main track and the arm track. There is a always a separate motor to rotate the arm.

BTW 3.5kg force and am using 21cm wheel and torque is 38kg which am using

I remember u said abt idealr wheels, what if i drive them them also by 20 rpm 4 kgcm torque motor.. would it help me drive the track?

38kg is not torque. 38kg is a force. If you meant 38kg-cm then that is torque. A wheel with a radius of 11.5cm would produce 3.3kg of force from 38kg-cm of torque:

Torque = Force * Radius
38kg-cm = 3.3kg * (21cm/2) = 3.3kg * 11.5cm

You NEED to know how much your robot weighs to figure out the torque you need for it to move and climb. (You also have to know your wheel radius or drive sprocket radius too)

For tracks...and any climbing robot you don't want it to go fast since it means gearing for more speed and less torque and it might fly over a bump or skid too much because it is trying to climb too fast. I'd go for as much torque and low speed as you can find. Larger wheels especially need this because larger wheels climb better but need much more torque and lower RPM because they have such a long lever arm and can easily go to fast because of their large circumference. It's not as big a problem with tracks since you can just use very large leading pulleys (or a shaped track) for climbing ability and then use very small drive sprockets to get high force and low speed for the a give amount of torque and RPM from the motor.

 

[dknguyen]

To build what is drawn in your design on the right takes the same amount of work and parts (plus a little more because of more motors) than the I-Robot designs that share the same shaft. In both cases you still need a track that can rotate independently of the arm which means you need a hollow shaft, except in your design you have more drive motors and need a longer arm that needs a higher torque servo which also has less usable length because the pivot point positioned farther back than the leadign edge of the robot.

Not sharing the same shaft as the pulley/sprocket for the main track has a lot of disadvantages. You have a longer arm so need a higher torque servo. There is less usable length in the arm to climb because it's pivot point is farther behind the leading edge. You need a separate drive motor for teh arm tracks and the main track on each side (4 drive motors) and unused torque can't be transferred between the arm track and main track because of separate drive motors.

It is a rather curious mechanism isnt it that can rotate the wheel to control the angle of the arm but have the track on teh arm spin without spinning the arm? This is the only way I know to pull this off:

-the same motor and driveshaft is used to spin the main track and the arm tracks (no separate motors for teh main tracks and arm tracks to maintain equal torque and speed and provide full torque to any part of the track that needs it)
-the drive sprocket for the main track and the drive sprocket for the arms are rigidly mounted to the same shaft that they share (duh!)
-the arm's frame itself sits on the drive shaft using a bearing so it can spin around teh shaft independently of the drive sprockets (a little less obvious unless you think about how it must work)
-the idler pulleys are mounted on the arm's frame through bearings so that they can free-spin (obviously)
-the driveshaft is HOLLOW. The outer driveshaft is used for driving the tracks and the arm's frame spins freely around this shaft via a bearing aha! this is where it gets fancy
-the inner shaft is RIGIDLY connected to the arm's frame and connects it to the drive servo and this controls the angle of the arm and that's the last piece of the puzzle

Is that a clear description for you?

 

[aswincool]

That Pdf is a great work, thanx

It is 38 kg cm torque, and the m/r rating is same,
I calculated it as 3.5 * 10.5 = 36.75 Kg cm
The above calculation was done for the model which i proposed earlier, in which only one wheel is attached to each motor on the front side, but now as per the design[best ever ] you proposed am going to drive 2 wheels in a common shaft, does this affect torque by any means?

 

[dknguyen]

In case you didnt realize, you can can drive the track from elsewhere. You dont need to drive the outershaft directly with the motor. The track will transfer power wherever it is put into the belt over to the outershaft similar to a belt or chain drive system (just make sure you use a sprocket instead of a idler pulley at the front since an the main track wont be able to transfer power to the arm track using an idler pulley and make sure the sprocket is the same size as the drive sprocket or you will get the arms trying to move at a different speed than the main track). It makes things easier since if you think about it, you can't just connect a motor onto the end of the outershaft to turn it like you normall would because the inner shaft comes out there and connects to the servo. It might make thigns very cramped with the servo and drive motors all being at the front. You have to drive the outershaft from the side whcih means either a small gear on the outershaft or a belt/chain drive system...but you already have on because you have the track. You could place the drive motors and drive sprocket at the back or anywhere else where there is mroe room and mount it to a regular solid shaft.

The required torque is only affected by the sprocekts being driven by the motor and the sprockets that are driving other belts (like the one that drives the arm track and is beign driven by the main track). The size of the other sprockets and idler pulleys don't matter since they just go along for the ride. Hollow shaft doesnt change anything.

You mention 3.8kg for soem wheels and 2kg for other wheels. Thsi doesnt make sense because of what I just said. A motor driving a shaft spins the wheel using torque because its spinning the wheel at the center. A driven track turning other sprockets uses FORCE because it pushes the sprocket from the edge, not rotating its center. It's like pulling on a rope and dragging along thigns tied along the rope. They all have the same force applied to them.

If you use the belt to drive a sprocket and have a shaft at the center of the sprocket to turn something else, then that's torque. Just like pulling a rope to turn a pulley attached to a shaft.

What havign more track does affect is the pressure on the ground. That's why tracked vehicles can move over softer ground than wheels. The total friction is the same, except it's spread thinly over a larger area rather than being concentrated in a small spot. Friction is independent of area in case you didnt know. It only depends on the normal force.

Also, torque is actually m*r, not m/r. Kg-cm, not kg/cm.

 

[aswincool]

sorry dude that's short form for motor [m/r]

 

[aswincool]

"Friction is independent of area in case you didnt know." so in some design links which u gave, usage of wider belt and narrow belt makes no difference? Surface area in contact makes no difference in grip??

If i use a belt of 2cm wide or 4cm wide in my design it makes no difference?

 

[dknguyen]

Oh okay lol. I thought you were using it as mass-radius like torque.

Thee width of the belt makes no difference in the friction. It reduces the pressure on the ground so it won't sink into soft soil or snow or mud as easily, but total friction is the same (which is often pretty significant). It's pretty hard to get your car stuck in snow if it doesn't sink in. So you wont have extra traction but you wont bog down and get stick like that.

When you increase the contact patch you increase the total area and the force is spread more thinly over that area. Decreasing the contact patch decreases teh total area but the force is more concentrated in that smaller area (ie. more pressure).

 

[aswincool]

Hmm ok i now understand the theory part of it, when width inc. force is distributed
But am not in a idea of using a sprocket till now.. Am using just wheels and belt, i was suggested to use sprocket since there would be slippage between the wheel's gripped OD and smooth inner side of the belt.. If this is the case should the belt be wider? so that it can gain more contact surface with the wheel ?

I thought if idler wheels are driven by a motor it will also add some pulling force in belt.. seems am wrong its k

The design tread arms.pdf is awesome and it's possible with the usage of gear wheels to connect the shaft and motor, am searching for those items but i have made a back up design too will upload it soon Thanx for ur spl interests

 

[aswincool]

Am switching to this design only because gears at required sizes are not readily available in my locality..
Other than the 360 degree rotation constraint, are there any other issues in this design

 

[dknguyen]

Hmm ok i now understand the theory part of it, when width inc. force is distributed
But am not in a idea of using a sprocket till now.. Am using just wheels and belt, i was suggested to use sprocket since there would be slippage between the wheel's gripped OD and smooth inner side of the belt.. If this is the case should the belt be wider? so that it can gain more contact surface with the wheel ?

Yeah, you want to use sprockets to drive the belt because of slippage. It also makes it so you dont need to tension the belt as much, otherwise the belt will have to be very tight. A wide belt with more contact area might help not slip if the belt is tight. But it doesnt help if the belt is loose and part of the belt will always be loose.

Look at a bicycle chain, notice that one side is loose and the other side is tight. If you reverse direction then the side that is loose and under tension switches. The sprocket will either pull the belt into tension or push the belt so it is loose depending on the direction it is rotating and whether the drive sprocket is driving the belt from the middle, front or at the back. This is not good if you are using pulleys since when the sprocket pushes teh belt it gets a bit more loose and you lose tension and the belt might even pop off.

For example a track going forward with the drive sprocket at the front will pull the track on top so it is tight and push the belt on the bottom so it is loose. If you go backwards, then the top gets pushed and becomes loose. This is bad because it means that the loose belt on the bottom well flex over the ground and you lose traction and the belt might even pop off the sprocket. This is why you use idler pulleys to press the bottom of the belt onto the ground and to stop a loose belt from popping off.

If you have the drive sprocket at teh back then the bottom is pulled tight and the top is pushed loose when going forward. And the bottom is pushed loose while teh top is puled tight when going backwards.

If the drive sprocket is driving the just the top or the bottom of the belt from the middle, then one side (left or right) will be pushed loose and the other side will be pulled tight. If you have a sprocket in the middle of the belt that is driving the belt from the top and bottom at same time, then you have top-left, top-right, bottom-left, and bottom-right.

Some designs have the motor/drive sprocket at the front on one side and at teh back on the otehr side because of limited space for the motor. In this case the belt touchign the ground is always loose on one side and tight on the other.

You can use pulleys just to make sure the middle of the belt is pressed to the ground so it doesnt flex over bumps and if the pulley also has a lip then it also helps so the belt doesn't pop off which can be very important especially for long loose tracks. Of course if you are using toothed/timing belts, you can just use idler sprockets too and that stops it from flexing over the ground and popping off. Look at tank tracks. The two large sprockets at the front- one is the drive sprocket. And then there area ton of little pulleys in the middle. These are there to help keep the track from popping off and to press the track against the ground when it is loose. It's one reason why good tracks are very heavy and have lots of little parts.

That's why timing/toothed belts sprockets are preferred over toothless belts for tracks.

I thought if idler wheels are driven by a motor it will also add some pulling force in belt.. seems am wrong its k

Think about it. The idler is not driving the belt. The motor IS driving the idler but it is using the belt to drive the idler so the idler cant possibly be driving the belt. So no extra force.

 

[dknguyen]

Am switching to this design only because gears at required sizes are not readily available in my locality..
Other than the 360 degree rotation constraint, are there any other issues in this design

So you're going with 4 separate drive motors for the two main tracks and arm tracks? I don't see the point of that personally because you need all the same parts plus 2 extra drive motors. I dont understand why gear size availabilty would affect that either since the drive sprocket for the arm track and main track should be of the same size and you can always drive the main track using a regular sprocket, solid shaft, and drive sprocket at the rear. Then use the main track to drive the front sprocket which drives the hollow shaft which turns the drive sprocket for the arm track. You dont need to use gears at all and its easier, and cheaper if you dont.

You need a hollow shaft either way to be able to rotate the arms and drive teh arm track and that's the hardest part. The method to drive the hollow shaft from anwyhere is already there because you have the main track which can be used as a belt drive.

Or is the problem that you cant get a hollow shaft? Then yeah you have to rotate the arms with a servo motor lever rather than a servo motor shaft and you have limited rotation. You need a lever or two-joint linkage because the arm is pivoting around its drive sprocket but the servo motor is not driving the arm from that stationary pivot point. Instead the contact point between arm and servo motor travels in an arc. Thats the issue I see and you have to be a bit more careful about designign it so things dont jam up.

 

[aswincool]

Thanx dude From monday i will be able to experiment my designs, the first one am going to test is irobot design which u proposed me with main drive at rear end wheels and servo drive at the support arm Will contact u soon with the possible difficulties