How can you make turning smoother?

[roboticsx]

does anyone know how to make turning smoother? our robot movement is too jerky...

[darkember]

What kind of drive train are you running?

[EricDrost]

Run a 6 or 8 wheel tank drive with the center wheel(s) dropped 1/8".

[FrankJ]

Run mechanum wheels?

Any kind of tank drive with grippy wheels or tracks are going to have issues turning smoothly. It is because you have to slide the wheels side way. There are compensations: short wheel base, six wheel drive with a dropped center wheel, etc. Having enough torque is also critical. Either by two cims per side or low enough gearing.

[Akash Rastogi]

does anyone know how to make turning smoother? our robot movement is too jerky...

Please post a picture of your drivetrain so we can help you better. :)

[Alan Anderson]

Chris Hibner's white paper (http://www.chiefdelphi.com/media/papers/1443) should be required reading for anyone worried about smooth turning.

[stveception]

does anyone know how to make turning smoother? our robot movement is too jerky...

Square the inputs from your joystick into your robot drive. This will make the joystick less sensitive.

[dtengineering]

Hey... one of the new Calgary teams! Welcome aboard, and welcome to Chief Delphi. I'm hoping to get out to Calgary this year, and look forward to seeing your machine.

We had this problem with our first robot, a four wheel drive machine... it worked great on our cement and linoleum floors at school, but when we finally got to Toronto and put it on carpet... we couldn't turn worth a darn.

The problem is inherent to skid steering... the front and rear of the robot needs to skid sideways, in order for the robot to turn. Since we design the robots to have high traction wheels (which were great for pushing) they were awful for turning. We had to reduce the coefficient of friction of one set of wheels so that they could slide sideways.

With a six or eight wheel drive (which we eventually switched to) the trick is to lower the centre wheels... this reduces the "normal force" (hey, physics in action) on the front and rear wheels, making it easier for them to skid sideways.

If you are locked into a four wheel drive design, you might want to consider some omni wheels... http://www.andymark.com/Omni-s/51.htm they also work great with six and eight wheel drive machines.

Good luck! Hope you get the machine running well soon.

Jason

[Jibri Wright]

Hey... one of the new Calgary teams! Welcome aboard, and welcome to Chief Delphi. I'm hoping to get out to Calgary this year, and look forward to seeing your machine.

We had this problem with our first robot, a four wheel drive machine... it worked great on our cement and linoleum floors at school, but when we finally got to Toronto and put it on carpet... we couldn't turn worth a darn.

The problem is inherent to skid steering... the front and rear of the robot needs to skid sideways, in order for the robot to turn. Since we design the robots to have high traction wheels (which were great for pushing) they were awful for turning. We had to reduce the coefficient of friction of one set of wheels so that they could slide sideways.

With a six or eight wheel drive (which we eventually switched to) the trick is to lower the centre wheels... this reduces the "normal force" (hey, physics in action) on the front and rear wheels, making it easier for them to skid sideways.

If you are locked into a four wheel drive design, you might want to consider some omni wheels... http://www.andymark.com/Omni-s/51.htm they also work great with six and eight wheel drive machines.
Good luck! Hope you get the machine running well soon.

Jason

It doesnt cause the normal force to decrease. What happens is that the centripetal force when pivoting on the dropped wheels causes the outer wheels to lift and cause them to be in a state of balance with each other. While pivoting the outer wheels are applying no force to the ground or in other words not touching.

[Ether]

With a six or eight wheel drive (which we eventually switched to) the trick is to lower the centre wheels... this reduces the "normal force" (hey, physics in action) on the front and rear wheels, making it easier for them to skid sideways.

It doesnt cause the normal force to decrease.

Um... yes it does.

What happens is that the centripetal force when pivoting on the dropped wheels causes the outer wheels to lift

Where are you getting your information?

[BJC]

does anyone know how to make turning smoother? our robot movement is too jerky...

To interpret this a different way, if you're talking about the robot turning really suddenly and being hard to control, make sure it weighs the full amount. Light weight robots are often squirrely. Also running on carpet vesus tile makes a huge difference.

If the robot IS having the opposite problem ( it's difficult to turn) and you are running a 6/8 wheel drive you can do a couple of things.

1. Change the outer wheels on one/both sides to a lower coefficient wheel. In general FRC terms Omni wheels have the least CoF (aside from Lunacy wheels [ew]) and pneumatic wheels have the most. The old gray rubber wheels were a happy medium between the two. Wedgetop is also a pretty good inbetween after it's been run on for a while.

2. Increase the drop in your chassis. If it is impossible to actually lower the middle wheels with your current design. turning down your outer wheels could work (although less desirable because it changes the surface speed of those wheels relative to the middle.)

3. CoG and tortional stiffness. It's probably too late to change these but your frame stiffness and keeping the robot's Center of Gravity close to the ground are large factors in a high performing chassis.

Regards, Bryan

[jar24]

Hey... one of the new Calgary teams! Welcome aboard, and welcome to Chief Delphi. I'm hoping to get out to Calgary this year, and look forward to seeing your machine.

We had this problem with our first robot, a four wheel drive machine... it worked great on our cement and linoleum floors at school, but when we finally got to Toronto and put it on carpet... we couldn't turn worth a darn.

The problem is inherent to skid steering... the front and rear of the robot needs to skid sideways, in order for the robot to turn. Since we design the robots to have high traction wheels (which were great for pushing) they were awful for turning. We had to reduce the coefficient of friction of one set of wheels so that they could slide sideways.

With a six or eight wheel drive (which we eventually switched to) the trick is to lower the centre wheels... this reduces the "normal force" (hey, physics in action) on the front and rear wheels, making it easier for them to skid sideways.

If you are locked into a four wheel drive design, you might want to consider some omni wheels... http://www.andymark.com/Omni-s/51.htm they also work great with six and eight wheel drive machines.

Good luck! Hope you get the machine running well soon.

Jason

Instead of the omni wheels from andymark I suggest you build your own, we learned the omni wheels they sell are kinda week if you get another robot pushing on you and your trying to maneuver out or away, My team has built our own omni and mecanum wheels for a few years now and they work just as good if not better and are a lot better on the teams wallet. Good luck, and hope to see you at worlds

[dtengineering]

Instead of the omni wheels from andymark I suggest you build your own...

The omni wheels from AndyMark work brilliantly and are used by hundreds of FRC teams... but I'm glad there are teams out there who choose to build their own, even if only for the fun of doing so.

But there is a pretty good chance your turning problems can be solved with a quick visit to Andy's website if you don't feel like machining a new part, from scratch, with a week left to go. Just order the correct diameter omni wheel that matches your current hub and pay for the expedited delivery and chances are you'll be turning smoothly by Wednesday.

Jason

[Al Skierkiewicz]

robiticxs,
What you describe is classic for four wheel drive systems as described above. What you don't see is the current drawn in this configuration. Current skyrockets as the drive motors consume near stall currents turning. Often the jerking occurs when the wheels finally break free of the carpet for a moment and then grab again. Omni wheels do correct this when added to the front or the back, not both, please. There are other solutions. Teams have put down a turning wheel that is at right angles to the drive wheels. When they want to turn, the additional wheel descends using pneumatics or some other power and lifts two of the wheels off the ground. A simple pneumatic cylinder with a caster wheel or nylon tip works almost as well.

[Tem1514 Mentor]

Al just a question when you use a drop wheel to aid in turning.

Do your bumpers have to stay within the bumper zone?

As I read and try to understand the rules, I would say Yes you do, but could be wrong.

[FrankJ]

Not trying to be Al, but yes your bumpers have to stay in the bumper zone.

You do not have to lift much to get the desired effect.

[Al Skierkiewicz]

The bumpers must stay within the bumper zone when the robot is flat on the floor.
4.1.6.4 R25
BUMPERS must be located entirely within the BUMPER ZONE, which is between 2 and 10 in. from the floor, in reference to the ROBOT standing normally on a flat floor.
Any robot using the six wheel (or more) dropped center, will rock and so will the bumpers. When climbing,pushing or interacting with the pyramid, the robot is expected to change attitude ans so too the bumpers. Also bumpers do not need to be parallel to the floor if that helps in your design.
Does that help?

[Jibri Wright]

Um... yes it does.



Where are you getting your information?




It doesn't cause normal force to technically decrease. When driving forward or backwards, there are at least four wheels on the ground. During this time, the amount of normal force on the dropped wheels as well as the normal force on the set of outer wheels is evenly distributed and the same. When the robot is pivoting on the dropped wheels, the normal force pushes in tangential to the place of pivot, this case being the dropped wheels. Inertia therefore causes the outer wheels to lift and balance which in turn means the dropped wheels are the only wheels touching the floor at this moment. I got this information from A.P. Physics B where I am currently getting a high B in. I learned of centripetal force last year and asked my mentor who also happens to be my A.P. Physics teacher is this what caused a dropped 6wd to turn smoothly. He said yes.
Study centripetal force and you'll know what im talking about.

[Ether]

It doesn't cause normal force to technically decrease.

You need to go back and read Jason's post more carefully:

this reduces the "normal force" (hey, physics in action) on the front and rear wheels

Dropping the center wheels has exactly the effect described above by Jason, and this is the main reason why dropping the center wheels slightly makes the robot easier to turn.


I learned of centripetal force last year and asked my mentor who also happens to be my A.P. Physics teacher is this what caused a dropped 6wd to turn smoothly. He said yes.
Study centripetal force and you'll know what im talking about.

Could you please politely ask him to send me a PM?

[Jibri Wright]

You need to go back and read Jason's post more carefully:



Dropping the center wheels has exactly the effect described above by Jason, and this is the main reason why dropping the center wheels slightly makes the robot easier to turn.




Could you please politely ask him to send me a PM?




Sure but he isn't on this website. Can you PM me your email? That way he can get in touch with you. He's the one that founded our robotics team and not really an engineer, but helps us determine how things should theoretically behave.

[TheOtherGuy]

Sure but he isn't on this website. Can you PM me your email? That way he can get in touch with you. He's the one that founded our robotics team and not really an engineer, but helps us determine how things should theoretically behave.

Jibri, what Ether is saying is that when you have weight centered on the robot and you add two dropped wheels, the weight is more distributed over the center wheels than the outer wheels, causing a higher normal force on the center wheels than either the front or back set of wheels. Your point about the centripetal force is still valid though.

[Jibri Wright]

I just talked to him. He said that the centripetal force does cause the outer two wheels to lift. Its the same force that NASA is currently studying to create "artificial gravity" in space stations. Centripetal force (in the form of static frictional force in this case) pulls inward and the outer wheels are pulled out parallel two the floor. As the robot begins to pivot, the centripetal force is out of balance. If you have ever done work using vectors, you can tell that one of the outer wheels on the ground and one wheel floating is what causes this imbalance. The centripetal force in turn forces a balance by making the forces on the two outer wheels equal a.k.a. set their vectors parallel to each other.

Think of it like a spinning top. When the top isn't spun, it is laying sideways on the table right? But when you spin it, it balances on a single point that touches the table, almost as if defying gravity. The force that is causing the top to balance is the same force that is causing the outer two wheels to lift.

My mentor also said that this observation is not as observable with an FRC 6wd because the dropped wheel is not dropped enough to notice, the FRC field carpet is too high to notice, FRC robot weight normally isn't evenly distributed on both sides of the robot, etc. so its hard to notice and may not come fully off the ground, but the wheels do lift.

He also said that this only occurs when the robot is pivoting. When the robot is driving forwards or backwards, four wheels are normally touching the ground and the weight (or as you said normal force) is even on all four. One set of wheels will be off of the ground and none of the weight will be on those wheels.

To put it plainly, since the wheels are probably touching the carpet when pivoting, normal force is decreased on the outer wheels when pivoting but not because of the dropped wheel. It is because of the centripetal force acting on the robot by tangential friction. Because a lot of Newtonian mechanics use a lot of the same measurements like Newtons, seconds, meters, etc.., it isn't always easy to figure out exactly what causes what. Even so in this case, it is centripetal force. The dead giveaway is the circular motion.

Btw Ether, love the Einstein picture. I love how he's the father of quantum mechanics yet dedicated his life to proving it wrong!

[Jeffy]

I just talked to him. He said that the centripetal force does cause the outer two wheels to lift. Its the same force that NASA is currently studying to create "artificial gravity" in space stations. Centripetal force (in the form of static frictional force in this case) pulls inward and the outer wheels are pulled out parallel two the floor. As the robot begins to pivot, the centripetal force is out of balance. If you have ever done work using vectors, you can tell that one of the outer wheels on the ground and one wheel floating is what causes this imbalance. The centripetal force in turn forces a balance by making the forces on the two outer wheels equal a.k.a. set their vectors parallel to each other.

Think of it like a spinning top. When the top isn't spun, it is laying sideways on the table right? But when you spin it, it balances on a single point that touches the table, almost as if defying gravity. The force that is causing the top to balance is the same force that is causing the outer two wheels to lift.

My mentor also said that this observation is not as observable with an FRC 6wd because the dropped wheel is not dropped enough to notice, the FRC field carpet is too high to notice, FRC robot weight normally isn't evenly distributed on both sides of the robot, etc. so its hard to notice and may not come fully off the ground, but the wheels do lift.

He also said that this only occurs when the robot is pivoting. When the robot is driving forwards or backwards, four wheels are normally touching the ground and the weight (or as you said normal force) is even on all four. One set of wheels will be off of the ground and none of the weight will be on those wheels.

To put it plainly, since the wheels are probably touching the carpet when pivoting, normal force is decreased on the outer wheels when pivoting but not because of the dropped wheel. It is because of the centripetal force acting on the robot by tangential friction. Because a lot of Newtonian mechanics use a lot of the same measurements like Newtons, seconds, meters, etc.., it isn't always easy to figure out exactly what causes what. Even so in this case, it is centripetal force. The dead giveaway is the circular motion.

Btw Ether, love the Einstein picture. I love how he's the father of quantum mechanics yet dedicated his life to proving it wrong!

While centripetal force exists, I'm not sure it does a lot for helping you turn on an FRC robot.
Also a test to show centripetal force is causing the better turning (or isn't) would be to decrease velocity of the turns and see if you notice more "scrubbing" due to a lower centripetal acceleration.

Also, I don't see a way that centripetal force/acceleration could be shown to be the reason an eight wheel drive with four dropped center wheels works.

Do a little search for "wheel base" on here and I'm sure you can find the reason drop center 6wd/8wd works the way it does.

[TheOtherGuy]

I just talked to him. He said that the centripetal force does cause the outer two wheels to lift. Its the same force that NASA is currently studying to create "artificial gravity" in space stations. Centripetal force (in the form of static frictional force in this case) pulls inward and the outer wheels are pulled out parallel two the floor. As the robot begins to pivot, the centripetal force is out of balance. If you have ever done work using vectors, you can tell that one of the outer wheels on the ground and one wheel floating is what causes this imbalance. The centripetal force in turn forces a balance by making the forces on the two outer wheels equal a.k.a. set their vectors parallel to each other.

Think of it like a spinning top. When the top isn't spun, it is laying sideways on the table right? But when you spin it, it balances on a single point that touches the table, almost as if defying gravity. The force that is causing the top to balance is the same force that is causing the outer two wheels to lift.

My mentor also said that this observation is not as observable with an FRC 6wd because the dropped wheel is not dropped enough to notice, the FRC field carpet is too high to notice, FRC robot weight normally isn't evenly distributed on both sides of the robot, etc. so its hard to notice and may not come fully off the ground, but the wheels do lift.

He also said that this only occurs when the robot is pivoting. When the robot is driving forwards or backwards, four wheels are normally touching the ground and the weight (or as you said normal force) is even on all four. One set of wheels will be off of the ground and none of the weight will be on those wheels.

To put it plainly, since the wheels are probably touching the carpet when pivoting, normal force is decreased on the outer wheels when pivoting but not because of the dropped wheel. It is because of the centripetal force acting on the robot by tangential friction. Because a lot of Newtonian mechanics use a lot of the same measurements like Newtons, seconds, meters, etc.., it isn't always easy to figure out exactly what causes what. Even so in this case, it is centripetal force. The dead giveaway is the circular motion.

Btw Ether, love the Einstein picture. I love how he's the father of quantum mechanics yet dedicated his life to proving it wrong!

The spinning top is an interesting example. What happens when the top is imbalanced and spinning slowly?

[Jibri Wright]

While centripetal force exists, I'm not sure it does a lot for helping you turn on an FRC robot.
Also a test to show centripetal force is causing the better turning (or isn't) would be to decrease velocity of the turns and see if you notice more "scrubbing" due to a lower centripetal acceleration.

Also, I don't see a way that centripetal force/acceleration could be shown to be the reason an eight wheel drive with four dropped center wheels works.

Do a little search for "wheel base" on here and I'm sure you can find the reason drop center 6wd/8wd works the way it does.

While decreasing the velocity, it would skid more.

Looking at equations, (Fc=centripetal force, ac=centripetal acceleration)
Fc=m*ac
ac=lvl^2/r
So looking at this as circular speed decreases, so does the acceleration and in turn the centripetal force. Less centripetal force means less gravity it can counteract and balance.

As for the 8wd, the reason that works well is because the wheels that are pivoted on are close to the pivot point of the robot. This means that the frictional force they provide produce less torque on the robot than wheels that are further away from the pivot point. An example is something you all probably have used which is an allen wrench. If you tighten a screw with it with the long way in the screw, it gets hard to turn. You then turn it to put the short way in and it gets easier to tighten. This is the same thing happening here.

[Jibri Wright]

The spinning top is an interesting example. What happens when the top is imbalanced and spinning slowly?

Centripetal force decreases due to friction on the top by the surface and maybe some air resistance. Eventually, gravity begins to start winning again and begins pulling a side of the top down.

[TheOtherGuy]

Let's assume you have a robot frame with electronics and the center of mass is directly in the middle of the robot, a few inches from the floor. Let's say it weighs 30lbs. Now, you add the battery (assume 15lbs), which is equidistant from the two center wheels and the two rear wheels.

Let's say you're spinning it slowly such that centripetal force is negligible. There is more normal force on the center wheels than the outer wheels because the center of mass is somewhere between the battery and the center of the robot (and more toward the center at that). This means there is less frictional force applied to the outer wheels and they will effectively "skid" across the floor, allowing you to turn on a dime.

If you increase the turning rate and take centripetal force into account, you are applying a force on the frame that is opposing the centripetal force on the battery.

Imagine you have a heavy weight on the end of a string that is attached to a vertically mounted pole. When you swing the weight around the pole (think tetherball), the weight indeed rises. However, the robot isn't the weight in this scenario. The robot is the entire system, and the weight is the center of mass. The pole tends to lean in the direction of the weight because of the force opposite centripetal force that is being applied to it. The same thing happens for FRC robots.

I believe it's much less noticeable because the dynamic friction is less than static friction in the case of wheel tread and with a robot weight mostly centered but off to one side, the center wheels will never start slipping. I have seen Vex robots with six wheel drive behave like four wheel drive robots, because the center of mass is either equidistant from the four wheels in contact with the floor or more toward the outer wheels.

[android1254]

We had that issues in our 2008 season overdrive game. The way we fixed it was getting omni
Wheels and kiddie corner them from each other so we could turn at full speed and it worked be very good. But before we got them in we had a comp so what we did was zip tied the the front left wheel and back right wheel so it have the same effect as the omni wheel would have. I hope this helps and good luck.

[Al Skierkiewicz]

Jibri,
While centripetal force may exist (on an FRC robot), I do not believe it is the dominant force in turning at the velocity we encounter on the field. Simply acceleration or deceleration cause the robot to tip towards the front or back wheels and generally halves the wheelbase in doing so. Teams have found that the distance the center wheel is dropped has significant effects on turning friction dependent on several factors such as wheel tread width, tread type and robot weight. In these designs, it is arguable how much weight is transferred to each wheel but it is unlikely that each wheel will ever receive equal force. In rare cases where the team has managed to balance the robot weight over the center wheels and keep the COG low, it is possible for all the weight to rest on the center wheels. This leaves the front and back wheels (in a six wheel robot) to act more as training wheels than actually supporting the robot. In these robots, turning is accomplished almost exclusively on two wheels minimizing the friction as discussed above. High speed, arcing turns also bring other forces into play shifting significant robot weight to the outer wheels.

[thmeans06]

Are you using victors? If so, make sure that the dynamic brake is turned off. If it isn't, when you let go of control your robot will stop abruptly.

[Jrat47]

You can use Omni wheels but you might lose some pushing power.