Optimal roller distance for dynamometer

[yawmid]

TL;DR Summary

building a dyno, need to determine the best roller distance

I am currently building a dynamometer to measure the power consumption and rpm of a mars prototype. The plan is to place each wheel of the rover onto a pair of rollers but I am currently struggling with determining the optimal roller separation. If they're too far apart, the wheel might not fully rotate the rollers and if they're too close they might roll over the top of the rollers.

 

[berkeman]

Welcome to PF.

What research have you done on this problem? Dynos are commonplace, so you should be able to see how others have solved this problem.

Also, can you show us your Free Body Diagram (FBD) for one of the wheels on the rollers and show all forces and torques involved? What is the range of weights and wheel torques that you have to accommodate?

Finally, you may want to rethink the mechanism that you have drawn. You probably need a bit of space between the wheel and the side guards, and some way to keep the vehicle and wheels tracking straight so they don't try to climb up and off of the dyno.

 

[yawmid]

Thank you for replying,

In terms of research I haven't seen anything that discusses this specifically.

I'm not quite sure how to approach the problem, below is the FBD diagram I have so far. I don't know the torque of the motor being used. The max weight on a roller pair will be around 120N.

You make a good point about the space required for the wheel and I'll be sure fix that. Keeping the vehicle straight is not an issue, I should have described the design in more detail which I will do now.

We are building a test rig for our prototype rover which needs to simulate various terrains and measure the power consumption/rpm of each of the 6 wheels. To do this, we are building essentially building dynos on top of scissor lifts (which does sound absurd without context). The scissor lifts will be able to move up and down to mimic inclines. A more detailed CAD model of the set up is shown below.

 

[berkeman]

Thanks, that helps some. But I'm not sure relying on the separation of the rollers at each wheel will be enough to keep the vehicle from climbing off the setup. I did a Google search on how hold back vehicle on dyno and got lots of helpful hits. Have you considered other hold-back mechanisms like those?

Paging @jrmichler @Ranger Mike @Baluncore

 

[berkeman]

And just to add -- what range of accelerations and slopes are you planning on testing to? How are you going to generate the counter-torques / loads at the rollers to perform that testing?

 

[Baluncore]

You might increase the roller diameter, then place a slack loop of belt over the rollers, that will conform to and support the vehicle wheel.

 

[yawmid]

Thanks, that helps some. But I'm not sure relying on the separation of the rollers at each wheel will be enough to keep the vehicle from climbing off the setup. I did a Google search on how hold back vehicle on dyno and got lots of helpful hits. Have you considered other hold-back mechanisms like those?

Paging @jrmichler @Ranger Mike @Baluncore

we are looking into some of these options now, but I'm still concerned about the roller distance. We want the rollers to rotate equally, would the direction of the wheels rotation change this? i.e if the wheel rotated clockwise would one roller rotate quicker than the other and is there anyway of stopping this. Our design will be subjected to a design review in front of some of the academics of my university so we have to think about every design choice as they will question everything.

 

[yawmid]

And just to add -- what range of accelerations and slopes are you planning on testing to? How are you going to generate the counter-torques / loads at the rollers to perform that testing?

The max slope will be 45 degrees, the acceleration is hard to come by because I'm not sure which motor will drive the wheels. That being said I am sure that it won't be anything too drastic. I'm not quite sure what you mean by counter-torques

 

[yawmid]

You might increase the roller diameter, then place a slack loop of belt over the rollers, that will conform to and support the vehicle wheel.

yeah that's a nice idea, I might implement it

 

[jack action]

Might be of interest to you:

SINGLE-ROLLER CHASSIS DYNAMOMETERS​

Single-roller chassis dynamometers, as vehicle power increases, have over time become the most popular solution among all the different manufacturers of chassis dynos.

As we can see from the analysis below, tyres find their optimal operating conditions on the road (image A).

Looking at the simulations below (images B and C) we can observe how at the larger diameter of the roller, corresponds to less 'indentation' and thus less stress (image. B) for tyres, also obtaining:

• Increased grip (larger contact area)
• Improved maintenance of constant operating conditions, during testing

​

TWIN-ROLLER DYNO TEST BENCHESE​

Originally born for diagnostic needs (at a time in history when vehicles had much lower powers than today) offered good stability of the vehicle under test, but sacrificed tyre retention.

This design choice was at the time also justified by the purposes for which the double-roller bench was originally designed; or its use within diagnostic centres that they were merely doing short and quick power tests.

HIGH STRESS CONDITIONS FOR TYRES​

As can be seen from the figures below (images. D, E) the car or motorbike test bench with double rollers (whether connected by a belt or not) submit anyway tyres under high stress conditions, also due to high indentation due to theimpossibility of using large-diameter rollers, implicit in these solutions.

INCREASINGLY POWERFUL OF MODERN VEHICLES​

These stressful conditions become increasingly evident as the power and torque of vehicles increase.

We therefore think of the modern endothermic, hybrid or electric vehicles, the torque and power of which have grown disproportionately in recent years, reaching more than double levels compared to the powers for which this type of bench double rollers was initially put on the market (1980s).

Typically, conducting medium/long-term tests, tyre warm-ups are likely to affect the results; due to:

• different grip in tyre/roller contact at different temperatures (cause of energy dissipation and variation of wheel power readings and the power dissipated.
• tyre diameter variations, which could occur due to their increased internal pressure with possible differences in engine speed readings.

DIFFERENT WHEEL AND ROLLER SPEEDS​

Further criticality of double roller dyno test benches is represented by the different rotation speeds of tyres and rollers.

During the acceleration period, tyres naturally tend to want to 'go up' on the front roller (image D), deforming even more due to displacement and different vehicle mass distribution.

This also involves the displacement of wheel and roller centres of rotation (and relative rolling diameter), modifying rotation speeds accordingly of the individual components involved (front roller/rear roller/tyre).

For understand this dynamic, one only has to look at Fig. D and E below; the distances from the axes of rotation from the centre of rotation of the tyre, change as the position of the vehicle axle changes.

This creates two different possible operating conditions, depending on the presence or not a mechanical constraint between front and rear rollers:

• When NOT bound together (Image D), the rear roller rotates at a higher rotational speed, condition that can cause slippage and adversely affect tests
• When bound together (image E), the rear roller carries a rubbing effect on the tyre surface, due to its lower rotational speed in this case; thus increasing tyre wear and operating temperature.

 

[berkeman]

I'm not quite sure what you mean by counter-torques

The resistance that the roller applies to the wheel. Without some resistance, the roller just spins up no matter how much power the wheel applies...

 

[Dullard]

See post #10.

It's not clear (to me) exactly what you want to test. The (inferred) point of an 'entire vehicle' test stand (as you describe it) is to test the entire vehicle. This is NOT the intent of most dynos. Going to the trouble of adjusting vehicle attitude via scissor-lift while using rollers for tire contact seems pointless. If you want to test (just) the power plant, don't bother with the wheels/attitude; if you want your test to include legitimate data for the tires, the dyno interface should 'look' like as much like the intended operating terrain as possible. Maybe I'm confused.