Devising a test to catch instability

Secrets, tips, tools, design considerations, materials, the "science" behind it all, and other topics related to building the cars and semi-trucks.
User avatar
Vitamin K
Pine Head Legend
Pine Head Legend
Posts: 1087
Joined: Sat Apr 20, 2013 7:26 pm
Location: Derwood, MD
Contact:

Devising a test to catch instability

Post by Vitamin K » Sat Mar 21, 2015 10:21 am

Hypothetical thinking caps on again...

So, the see-saw between maximizing potential energy while maintaining a stable car is a tricky one to manage, to be sure. The further back your CoM, the more likely you are to invite the "waggle of death" as your car speeds down the track.

The obvious remedies are to make the car as well aligned as possible (so as to keep the rear end going straight), and to provide enough steer to keep the car glued to the rail, as any oscillation is going to open up the door to that loaded butt dance that we all hate.

The frustrating part, of course, is that there's no good way to test whether or not your car is stable enough with a given CoM. Thus, the further back your weight, the more risks you take. Of course, if you have a test track, you have the luxury of actually watching to see how the car runs down the stretch, and can add and decrease steer accordingly.

Most of us, though, don't have tracks, since they require a good amount of money and space to use and acquire. A tuning board is good to set the initial alignment, but it doesn't tell us whether or not the car is going to be stable enough at speed.

So, consider this possibility:

- A 10-12 foot length of track...single lane, with a stop section at the end.
- Some kind of mechanism to launch a car from the starting point, roughly approximate to the amount of energy a car will have coming off of the curve on a real track. Maybe this could be a slingshot, or one of those dual-wheel setups that hot wheels launchers employ.
- A deliberately introduced "bump" somewhere on the rail, to introduce instability, and force the car to attempt to correct itself.

The idea is to launch the car so it's moving at "real" speeds, have it hit the tiny bump, and then watch to see if it continues straight, or starts oscillating down the rest of the travel.

Would this possibly replicate the conditions that set up a death waggle in a car? I don't have any idea of how practical or feasible this actually is...I'm just thinking aloud and dumping my thoughts here on the forum. :)



Topspin.D
Master Pine Head
Master Pine Head
Posts: 183
Joined: Wed Jun 13, 2012 12:10 am
Location: Kirkland, Washington

Re: Devising a test to catch instability

Post by Topspin.D » Sun Mar 22, 2015 1:10 am

Good ideas... but we have a 24' test track we built. The track is to "traditional" cub scout PWD track specifications. We just don't have that last 8' of run-out.

Well... we can tell you that our cars didn't show signs of wiggles on the test track, but when we went to race on a 42' Best track... we didn't have enough steer. That last 8-10 feet we didn't have would've shown the issues with the cars.... I think you learn more from how the car behaves in the flats than you do from watching it on the drop.



User avatar
Vitamin K
Pine Head Legend
Pine Head Legend
Posts: 1087
Joined: Sat Apr 20, 2013 7:26 pm
Location: Derwood, MD
Contact:

Re: Devising a test to catch instability

Post by Vitamin K » Sun Mar 22, 2015 6:28 am

Topspin.D wrote:Good ideas... but we have a 24' test track we built. The track is to "traditional" cub scout PWD track specifications. We just don't have that last 8' of run-out.

Well... we can tell you that our cars didn't show signs of wiggles on the test track, but when we went to race on a 42' Best track... we didn't have enough steer. That last 8-10 feet we didn't have would've shown the issues with the cars.... I think you learn more from how the car behaves in the flats than you do from watching it on the drop.
Yeah, that's sort of what I'm aiming for with the setup. The track section would lay flat, not sloped, with some kind of mechanism to propel the car into it, as if it were coming off of the curved section and onto the flat.

I should probably watch some race videos and see if I can identify any common characteristics to cars that get the wiggles downtrack.



ngyoung
Master Pine Head
Master Pine Head
Posts: 234
Joined: Thu Apr 03, 2014 6:36 pm
Location: Eyota, Minnesota

Re: Devising a test to catch instability

Post by ngyoung » Sun Mar 22, 2015 8:42 am

That's definitely going to be a challenge. My first thoughts were to use a push plate that slide on a slot on the rail, instead of pushing you could try using a pulley system, or simply use a more aggressive slope angle. Each option has drawbacks that will effect how the car runs when trying to accelerate to top speed faster.



User avatar
sporty
Pine Head Legend
Pine Head Legend
Posts: 3355
Joined: Sun Feb 13, 2005 2:00 am
Location: rockfalls, Illinois

Re: Devising a test to catch instability

Post by sporty » Sun Mar 22, 2015 2:07 pm

Tall task at hand here.

I don't see a all the effort worth what the solution would tell you.

I don't have the time or interest for this one.

I think I'd find a way to make a track or get a used one. Run it outside if needed. When space is a issue.

You are getting into slow motion capture of car at fill speed.

The newer extra sensor, timer. Software system seems to get u closer to knowing what a car is doing during certain parts of the track. To help you find the trouble areas.

Just don't see the effort for all the added work to find. What you can do with a easier way.

I use to drive to places. That had a track. To test or tune. When the kids wanted to do well.

We were always able to find other scout packs racing. Before or after our race and asked if we could come run on the track or pay and enter if they had open class divisions.

We even mailed them to other fund raiser races. Just to get the times and info.

I think there's a lot of options.

To even borrow a track.

Especially with packs getting newer tracks. Sometimes they just give away the old one or sell cheap.

There's a lot of easier options. I'd try first.



User avatar
Vitamin K
Pine Head Legend
Pine Head Legend
Posts: 1087
Joined: Sat Apr 20, 2013 7:26 pm
Location: Derwood, MD
Contact:

Re: Devising a test to catch instability

Post by Vitamin K » Sun Mar 22, 2015 2:26 pm

sporty wrote:Tall task at hand here.

I don't see a all the effort worth what the solution would tell you.

I don't have the time or interest for this one.

I think I'd find a way to make a track or get a used one. Run it outside if needed. When space is a issue.

You are getting into slow motion capture of car at fill speed.

The newer extra sensor, timer. Software system seems to get u closer to knowing what a car is doing during certain parts of the track. To help you find the trouble areas.

Just don't see the effort for all the added work to find. What you can do with a easier way.

I use to drive to places. That had a track. To test or tune. When the kids wanted to do well.

We were always able to find other scout packs racing. Before or after our race and asked if we could come run on the track or pay and enter if they had open class divisions.

We even mailed them to other fund raiser races. Just to get the times and info.

I think there's a lot of options.

To even borrow a track.

Especially with packs getting newer tracks. Sometimes they just give away the old one or sell cheap.

There's a lot of easier options. I'd try first.
You might possibly be the voice of reality here. However, I will still burn some thought cycles thinking it through some more. I won't have the money or space for any kind of real track for the forseeable future. This could be considered something of a compromise...a "velocity tuning board", if you will. A regular tuning board lets you watch the behavior of the car at slow speeds to fine tune drift and alignment. This would let you see the behavior of the car at high speed.

I accept that this may be a totally worthless endeavor...but that never stopped me before! :mrgreen:

Might sketch out a concept for a launching sling-shot later tonight...if I can avoid the wife and her honey-do list, that is...



dpatrick911
Apprentice
Apprentice
Posts: 12
Joined: Sat Mar 21, 2015 6:25 am
Location: Flat Rock, MI

Re: Devising a test to catch instability

Post by dpatrick911 » Sun Mar 22, 2015 2:38 pm

Any ideas good or bad regarding using a treadmill? If you could suspend the car or hold the car at the COM you might be able to see what the car is doing at speed. My idea seems a little suspect. But it might give you some data.



User avatar
Vitamin K
Pine Head Legend
Pine Head Legend
Posts: 1087
Joined: Sat Apr 20, 2013 7:26 pm
Location: Derwood, MD
Contact:

Re: Devising a test to catch instability

Post by Vitamin K » Sun Mar 22, 2015 3:33 pm

dpatrick911 wrote:Any ideas good or bad regarding using a treadmill? If you could suspend the car or hold the car at the COM you might be able to see what the car is doing at speed. My idea seems a little suspect. But it might give you some data.
Shhhh, you can't go talking about treadmills! Next, you'll be talking about bent axles and then you'll get ridden out on a rail...!

Ahem.

But, seriously. I think the behavior of the car as it rides on/against the rail is actually pretty crucial too this diagnosis, so I'm not certain that the behavior on a treadmill would be useful here.

Also, since most commercial treadmills have textured surfaces, you'd have to build your own if you wanted a smooth surface to ride on.



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Devising a test to catch instability

Post by Stan Pope » Sun Mar 22, 2015 4:43 pm

Vitamin K wrote:
dpatrick911 wrote:Any ideas good or bad regarding using a treadmill? If you could suspend the car or hold the car at the COM you might be able to see what the car is doing at speed. My idea seems a little suspect. But it might give you some data.
Shhhh, you can't go talking about treadmills! Next, you'll be talking about bent axles and then you'll get ridden out on a rail...!

Ahem.

But, seriously. I think the behavior of the car as it rides on/against the rail is actually pretty crucial too this diagnosis, so I'm not certain that the behavior on a treadmill would be useful here.

Also, since most commercial treadmills have textured surfaces, you'd have to build your own if you wanted a smooth surface to ride on.
A "treadmill" approach is not all bad. The "running surface" (bed) need not be 3' wide and 6' long either... 4"X8" should be sufficient! Next, a rail needs to be attached ... it could be stationary or move with the "bed". Stationary would be much easier, but not a true analogy. Possible the most difficult is a mechanism that holds the car in place but does not inhibit any of the likely motions. A possibility includes a line attaching the front of the car to a fixed point well forward of the end of the bed. (A short line would provide too much lateral force and inhibit some of the instability you are trying to detect.)

Instead of using a "belt" you might construct a wood "cylinder" 4" wide with an attached center rail, with the car intended to ride the outside top of the cylinder. The radius must be large enough that the height of the rail at the center of a 5" chord not exceed 3/32". (I don't know what radius would be required and my spreadsheet program is not available on this backup computer ... maybe someone can run those numbers to compute the radius and see if it is feasible.)


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Vitamin K
Pine Head Legend
Pine Head Legend
Posts: 1087
Joined: Sat Apr 20, 2013 7:26 pm
Location: Derwood, MD
Contact:

Re: Devising a test to catch instability

Post by Vitamin K » Sun Mar 22, 2015 4:57 pm

Stan Pope wrote: A "treadmill" approach is not all bad. The "running surface" (bed) need not be 3' wide and 6' long either... 4"X8" should be sufficient! Next, a rail needs to be attached ... it could be stationary or move with the "bed". Stationary would be much easier, but not a true analogy. Possible the most difficult is a mechanism that holds the car in place but does not inhibit any of the likely motions. A possibility includes a line attaching the front of the car to a fixed point well forward of the end of the bed. (A short line would provide too much lateral force and inhibit some of the instability you are trying to detect.)

Instead of using a "belt" you might construct a wood "cylinder" 4" wide with an attached center rail, with the car intended to ride the outside top of the cylinder. The radius must be large enough that the height of the rail at the center of a 5" chord not exceed 3/32". (I don't know what radius would be required and my spreadsheet program is not available on this backup computer ... maybe someone can run those numbers to compute the radius and see if it is feasible.)
If you go here, you can see some thoughts on treadmill testing, along with some pitfalls, and a small battery-powered design the author made. He doesn't address the issue of the rail, though, so that's another challenge. (Search for 'treadmill' on the page)

I am still skeptical of being able to accurately replicate a track running surface with any kind of treadmill setup. Also, there's the issue of prolonged running wearing down bores.

As for the cylinder...well...I see a lot of potential issues regarding constructing the parts and mechanisms for something like that.



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Devising a test to catch instability

Post by Stan Pope » Sun Mar 22, 2015 5:59 pm

Vitamin K wrote:
Stan Pope wrote: A "treadmill" approach is not all bad. The "running surface" (bed) need not be 3' wide and 6' long either... 4"X8" should be sufficient! Next, a rail needs to be attached ... it could be stationary or move with the "bed". Stationary would be much easier, but not a true analogy. Possible the most difficult is a mechanism that holds the car in place but does not inhibit any of the likely motions. A possibility includes a line attaching the front of the car to a fixed point well forward of the end of the bed. (A short line would provide too much lateral force and inhibit some of the instability you are trying to detect.)

Instead of using a "belt" you might construct a wood "cylinder" 4" wide with an attached center rail, with the car intended to ride the outside top of the cylinder. The radius must be large enough that the height of the rail at the center of a 5" chord not exceed 3/32". (I don't know what radius would be required and my spreadsheet program is not available on this backup computer ... maybe someone can run those numbers to compute the radius and see if it is feasible.)
If you go here, you can see some thoughts on treadmill testing, along with some pitfalls, and a small battery-powered design the author made. He doesn't address the issue of the rail, though, so that's another challenge. (Search for 'treadmill' on the page)

I am still skeptical of being able to accurately replicate a track running surface with any kind of treadmill setup. Also, there's the issue of prolonged running wearing down bores.

As for the cylinder...well...I see a lot of potential issues regarding constructing the parts and mechanisms for something like that.
Yes, his "treadmill" is lacking (and he asserts that it produces inferior alignment vs. conventional methods) and his alignment method is indirect, time consuming and destructive (i.e. can "produce" alignment, but not test alignment).

But we are looking for something to test a more specific trait ... rear end stability. (We already have alignment efficientlyl "nailed.") To do this, we need to simulate many real track features. I will offer (separately) some fabrication approaches that may simplify the creation of a drum that operates smoothly and accurately.


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Devising a test to catch instability

Post by Stan Pope » Sun Mar 22, 2015 7:48 pm

If I did the formulae right, the drum radius would have to be about 31.3" to allow a 5" wheelbase car easily clear a 1/4" rail. (I used 0.35" = 0.1" + 0.25") as the limit height of the rail between the wheels.)

Somebody please sanity check this!

If correct, the drum diameter would have to be about 5' ... way too big for my test environment! :)


Stan
"If it's not for the boys, it's for the birds!"

Laserman
Merchant
Merchant
Posts: 168
Joined: Sun May 05, 2013 8:21 am
Location: New York, NY

Re: Devising a test to catch instability

Post by Laserman » Sun Mar 22, 2015 8:32 pm

Hi Stan,
Why not have the car on the inside of the cylinder. Kind of like a hampster with a 4' diameter?
Getting that big cylinder spinning might take some precision parts. It would be amusing though.
Thanks



dpatrick911
Apprentice
Apprentice
Posts: 12
Joined: Sat Mar 21, 2015 6:25 am
Location: Flat Rock, MI

Re: Devising a test to catch instability

Post by dpatrick911 » Sun Mar 22, 2015 9:00 pm

My initial thought was a bar of teflon used as the center rail. Cutting down on wheel wear, of the DFW. Reasonable? Next was a pin sticking up thru the center rail. This pin would be fixed at the COM of your car or as close as your calculations can predict. You can adjust it from that starting point if necessary. But a rough approximation can be made that the COM should be located at some point along the longitudinal axis of the car. My thought is this. Hypothetically the COM should want to travel a straight line along the direction of travel, that being the center rail. Any rotation (yaw) should be generated with the COM being the axis of that rotation (yaw). Perhaps, there are too many other variables to take into account to get an useful data from this. But, the COM should always want to travel along the line of travel (following the center rail). Right? Just thinking out loud, honestly. I am probably the only person this makes sense to, there is likely a huge hole in this idea that I have not taken into account.



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Devising a test to catch instability

Post by Stan Pope » Mon Mar 23, 2015 1:51 am

"Hamster" -- when I considered that, I wondered how to hold the car in place. But, revisiting with your question, I see that gravity would do the job! :) In fabrication, I would probably machine the outside of the drum as though it were on a lathe. Machining the inside would be a bigger challenge. In any case, I forsee a lot of dovetail joints! :) Visibility and measurement could be difficult unless the drum could be supported with sufficient rigidity from just one side. (That scares me.)

"Pin through CM" -- Might work. Suppose the pin were free to move side to side to stay with the car and not constrain its sideways motion! I'm not sure where the pin should contact the car... CM might be the right place.

"Teflon" -- I'd really like the materials to be the same as on a real track, and I'd like the relative motion of the car and teh track parts to be the same. Somehow a moving track and a stationary rail grates on my sensibilities. :)

"Wear" -- I don't see very much time spent on this track by any one car ... just enough to see its stability as the track cycles down from 12 to 9 mph. So "wear" is probably not an issue.

I like the thinking that is happening, though! Don't stop, guys!


Stan
"If it's not for the boys, it's for the birds!"

Post Reply