LocostUSA.com

Learning how to build Lotus Seven replicas...together!
It is currently March 28, 2020, 7:26 pm

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 11 posts ] 
Author Message
PostPosted: October 2, 2019, 9:35 am 
Offline
User avatar

Joined: April 23, 2019, 4:30 am
Posts: 58
I have been contemplating the best way to test my cars dynamic handling on the limit of grip without turning a wheel, bear with me as I go through the theory.....



At any instant in time during dynamic cornering a car will be accelerated sideways at the contact patches by the centripetal acceleration, and the inertia of the vehicle acting through the center of mass (COM) will react through the suspension to balance all these offset forces. If you could apply a centrifugal force of 1G at the COM then you can simulate how the body reacts in 1G cornering whilst being static.

In a static test an understeering car will slip at the front when the COM side load overcomes the grip available at that axle, oversteering the rear will slip. The limits of grip should be easy to assess as the tire deflection, load transfer and suspension bush deflection are all "realistically" simulated. A digital weight scale on the fork connection would allow force readings to be recorded up to the point of skidding.

In my case my vehicle WILL (come hell or highwater) be 1300kg and I actually do have access through the frame to the COM position, so can mount an eyebolt and a wire rope to a forklift. By lifting a 1300kg "load" on the forklift the sheave wheel will allow me to apply side load of 1300kg simply. I think it will be possible to tune the suspension to slide simulateously at both axles then backoff the front grip to add a bit of understeer to the chassis.

This arrangement does not take into account dynamic factors such as aero loads, tire temperature or road surface variations. The entire setup is transportable so can be setup anywhere quickly and tire warmers could be pulled off just before the test and the vehicle rolled forward enough to put the hot sections at the contact patch. Extra weights can be added to simulate the aero load at each end if known (that is a whole other thread itself).

It is simple enough that I may just have to build one to try it out - once I have my car built! Therefore this thread won't be updated for a long time!

Attachment:
side load.jpg
side load.jpg [ 82.55 KiB | Viewed 1573 times ]


The sheave wheel frame is trapped under the front wheels of the forklift and therefore can resist the lifting force if sized appropriately - these models are purely conceptual. The vehicle will have to have its COM accurately measured then the wheel height is adjusted so the angle of pull is strictly horizontal - otherwise erroneous downforce or lift is added.
Attachment:
sheave wheel frame.jpg
sheave wheel frame.jpg [ 43.75 KiB | Viewed 1573 times ]


Disclaimer - Procedures for operating such an arrangement safely would be solely the responsibility of anyone who made such an arrangement - there are numerous hazards involved and this method is purely a theoretical concept only.

Cheers,
Marcus.


Last edited by Kinetic Research on November 30, 2019, 8:20 pm, edited 2 times in total.

Top
 Profile  
Reply with quote  
PostPosted: October 2, 2019, 12:10 pm 
Offline
User avatar

Joined: April 26, 2008, 6:06 pm
Posts: 3026
Location: Under the weather. (Seattle)
Even if you get it as accurate as possible, including driver, steering angle, and locking the rear axle, you're still only simulating one very limited scenario...Perfect steady state cornering. This would not account for any dynamic factors, nor driving style. So whatever you do, you'll still need to make adjustments during dynamic testing anyways. This means ending up in the same place either way. Assuming the same pre-testing measurements and math are performed initially to get the suspension in the right ballpark, you'll be starting in the same place as well. That's why I'm struggling to see significant value in this the extra time, effort, cost, and risk of performing this single scenario static testing in between. In order to have the exact COM, you need a completed car. If you have a completed car, why avoid turning wheels?

That being said, if you do see enough value in it, or simply want to because you want to...By all means go for it. If nothing else, it would be kind of interesting to see.

_________________
-Justin

Also follow my build on blogspot, tumblr, or instagram and twitter (GarageOdyssey)


Top
 Profile  
Reply with quote  
PostPosted: October 2, 2019, 2:24 pm 
Offline

Joined: January 31, 2008, 5:34 pm
Posts: 744
Location: SW Wes Consin
Where on a typical frame can you locate your 3D COM attachment ? And what about roll ?


Top
 Profile  
Reply with quote  
PostPosted: October 2, 2019, 3:01 pm 
Offline
User avatar

Joined: April 23, 2019, 4:30 am
Posts: 58
EDIT - doubled up post


Last edited by Kinetic Research on October 2, 2019, 5:24 pm, edited 1 time in total.

Top
 Profile  
Reply with quote  
PostPosted: October 2, 2019, 3:22 pm 
Offline
User avatar

Joined: April 23, 2019, 4:30 am
Posts: 58
vroom wrote:
Where on a typical frame can you locate your 3D COM attachment ? And what about roll ?


For a locost I don't know where your COM is located, anyone got pictures?

Given the midengine layout I am building the passenger rear bulkhead tubing is fairly much inline with the COM so a bracket could be easily added. The tubing in that area is strong enough to take the 1300kg side load as its also my main rollover protection.

If you apply a side force at the COM height the car will roll exactly like its cornering. The suspension will compress on the outside and load transfer will occur at the contact patches the same as if the car were actually cornering. The summation of all the suspension movement should (in theory) be no different than if you were doing it at speed, except you can measure with precision that instant - its like a freeze frame of the car at the limit (eureka - that's the name of this test contraption - the freeze frame!)

On the track centripetal acceleration and cornering is generated by turning the front wheels and the inertia of the vehicle - represented by the COM - wants to keep going straight. As long as the steering wheel is turned then the side force is still applied and the car stays in a rolled over attitude. Straighten up, the side load disappears and the car settles to its static ride height under 1g vertical loading. Bumps in the track affect the dynamics of all of this though....


Top
 Profile  
Reply with quote  
PostPosted: February 5, 2020, 4:12 pm 
Offline

Joined: January 2, 2020, 8:16 am
Posts: 3
I have an similar thought as you but somewhat realized in a different way:

Instead of a pulling jig (that would require a forklift in the presented case), I envisioned using the same device that FSAE judges use to evaluate the vehicle rollover potential: a tilting table. By tilting the car you can simulate a degree of lateral acceleration and if the car sits on 4 corner scales, one can evaluate the weight transfer and hence the roll moment distribution given the springs and anti-roll bars.

One limitation of my method is that the normal weight of the vehicle changes due to the angle and the projected lateral force but one can compensate by adding more weight to the car, I suppose.

Your idea would work well to, you only need the four scales to be able to get this very useful data.


Top
 Profile  
Reply with quote  
PostPosted: February 6, 2020, 1:34 pm 
Offline

Joined: July 17, 2006, 3:09 pm
Posts: 286
Location: Holly, MI
I think it will be interesting to see. My only thought is that one of the main things you are not simulating is the tire slip angle. The real world grip of a tire is very dependent on slip angle, and I believe you have to have the tires rolling for that. You also won't be seeing any of the longitudinal forces associated with drive torque or rolling resistance. I don't know how those forces change the contact pattern, but they do.

Ken


Top
 Profile  
Reply with quote  
PostPosted: February 6, 2020, 5:33 pm 
Offline
User avatar

Joined: April 23, 2019, 4:30 am
Posts: 58
BB69 wrote:
I think it will be interesting to see. My only thought is that one of the main things you are not simulating is the tire slip angle. The real world grip of a tire is very dependent on slip angle, and I believe you have to have the tires rolling for that. You also won't be seeing any of the longitudinal forces associated with drive torque or rolling resistance. I don't know how those forces change the contact pattern, but they do.

Ken


Thanks for the input Ken, not being a suspension or tire dynamicist I bet I am guilty of over simplifying. Slip angle, drive torque and rolling resistance are all critical to overall handling. https://www.youtube.com/watch?v=W8UiE7yvO_M

My rig is based around my understanding of the traction circle, hence is only valid for that micro instant were the car has neither drive nor braking vectors, but is purely side loaded - which would be the maximum point for cornering force. Its at this point where the car will understeer, oversteer or be neutral - any further control inputs like gas, brakes or steering will affect this balance point. I want to show some of your videos because they have influenced my thinking over the years (thanks for posting them, would love to see more!)

In the video (Picking through Grattan traffic) https://www.youtube.com/watch?v=5C0OgpvWuxc&t=9s the steering rarely turns more than 100 degrees off centerline. I don't know what ratio of steering wheel to wheel angle is in your setup, the Mclaren F1 runs 18.5 degrees steering to 1 degree wheel rate, so in that case the front end would only turn 5.5 degrees. The front wheel could be turned in the static testing to simulate this influence, then all geometry and contact patch pressures would also be replicated.

Front slip; the steering has to be turned to take a corner, and this affects the contact patch due to factors such as tire width and section, camber, caster, kingpin inclination, ackerman, tire vertical load, steering angle etc.

Rear slip: My rear end will be setup very close to yours - Wide back tires appear to mostly be deflecting; https://www.youtube.com/watch?v=2HEQDpLfncQ

Another way to look at the testing rig is if the load was applied purely from the front, it would simulate the maximum braking grip due to load transfer and geometry changes in the suspension (for discussion please ignore the critical effect of the disc friction vs temperature). The loading can only act through the COM to be "accurate".

First hurdle is to build my car first, the test track is 1.5 hrs drive away so I was thinking of ways to reduce my setup time and cost but its best done (and is more fun) on the track like everyone else does.

Its a pity the server apocolypse of 2019 wiped out the photos in your track monster build log - were you going to build a mark 2?

cheers,
Marcus.


Top
 Profile  
Reply with quote  
PostPosted: February 7, 2020, 5:00 pm 
Offline

Joined: July 17, 2006, 3:09 pm
Posts: 286
Location: Holly, MI
Kinetic Research wrote:
BB69 wrote:
I think it will be interesting to see. My only thought is that one of the main things you are not simulating is the tire slip angle. The real world grip of a tire is very dependent on slip angle, and I believe you have to have the tires rolling for that. You also won't be seeing any of the longitudinal forces associated with drive torque or rolling resistance. I don't know how those forces change the contact pattern, but they do.

Ken


Thanks for the input Ken, not being a suspension or tire dynamicist I bet I am guilty of over simplifying. Slip angle, drive torque and rolling resistance are all critical to overall handling. https://www.youtube.com/watch?v=W8UiE7yvO_M

My rig is based around my understanding of the traction circle, hence is only valid for that micro instant were the car has neither drive nor braking vectors, but is purely side loaded - which would be the maximum point for cornering force. Its at this point where the car will understeer, oversteer or be neutral - any further control inputs like gas, brakes or steering will affect this balance point. I want to show some of your videos because they have influenced my thinking over the years (thanks for posting them, would love to see more!)

In the video (Picking through Grattan traffic) https://www.youtube.com/watch?v=5C0OgpvWuxc&t=9s the steering rarely turns more than 100 degrees off centerline. I don't know what ratio of steering wheel to wheel angle is in your setup, the Mclaren F1 runs 18.5 degrees steering to 1 degree wheel rate, so in that case the front end would only turn 5.5 degrees. The front wheel could be turned in the static testing to simulate this influence, then all geometry and contact patch pressures would also be replicated.

Front slip; the steering has to be turned to take a corner, and this affects the contact patch due to factors such as tire width and section, camber, caster, kingpin inclination, ackerman, tire vertical load, steering angle etc.

Rear slip: My rear end will be setup very close to yours - Wide back tires appear to mostly be deflecting; https://www.youtube.com/watch?v=2HEQDpLfncQ

Another way to look at the testing rig is if the load was applied purely from the front, it would simulate the maximum braking grip due to load transfer and geometry changes in the suspension (for discussion please ignore the critical effect of the disc friction vs temperature). The loading can only act through the COM to be "accurate".

First hurdle is to build my car first, the test track is 1.5 hrs drive away so I was thinking of ways to reduce my setup time and cost but its best done (and is more fun) on the track like everyone else does.

Its a pity the server apocolypse of 2019 wiped out the photos in your track monster build log - were you going to build a mark 2?

cheers,
Marcus.


I'm no expert for sure, but that micro instant you are talking about is purely theoretical in my experience. The goal of maximizing the friction circle means you always have a longitudinal portion of acceleration. Also, under and over steer is a dynamic event. What I mean, is that your car will transition from one to the other rapidly. It's this transition you should really care about, not some instantaneous static point that is over before you can do anything about it. More than anything, you want a car that does the same thing lap after lap so you can learn what to do. If it understeers on entry and then rapidly changes to oversteer with power, you can learn to drive that way. If it understeers one lap, and then oversteers the next, you will always struggle to control the car and your lap times will suffer. Most importantly, you won't have confidence driving the car.


I have more Excel sheets than I can count, trying to calculate what the car is going to do. They are all pretty much useless for me when I'm at the track. There are a couple reasons for that, and I will explain those so you can decide if you will have the same issue.

1. Tires. You are talking about trying to tune the car around some admittedly small handling window. Are you willing and able to use tires in the exact same condition every time you make a change? I'm not. I use scrub tires from pretty much wherever I can get them. They are different sizes, brands, and mfg dates. This alone means I will never reach the ultimate in handling because I am constantly changing one of the biggest factors. If you go spend some time with the guys consistently winning races, you will see they are adamant about reducing variables when testing and racing. This means they go through a lot of tires. They will test a set for only a few laps, and then they will put on a new set for the next round of changes so they can eliminate the tires from the equation. This is anything but locost.

2. Attention to detail. There are lots of other factors that change the handling of the car, and I don't have the time or drive to control them every event or session. For instance, are you going to have the car aligned before every event? Small changes in toe or camber from a bushing moving or wearing will affect the car. Are you adjusting your tire pressures for track conditions every run? How are you making sure your differential bias is changing (big factor for handling assuming you are using a limited slip of some kind)?

In the end, wondering these things is very interesting for me. I'm working on a braking spreadsheet now to optimize my system. Once you have the car built and running though, you will realize you learn so much more for every lap turned compared to reading/calculating. So, build the rig in your spare time, but don't let it slow down the car build.

As for me, I'm always thinking about rev 2. I need a new garage first. In the meantime, I'm working on some custom rear control arms; inboard rear coilovers; power steering; ABS; and revisions to the overall brake system sizing. I always try to post more videos. I'm pretty sure I have all the photos from my build still, so ask if there is something you want to see.

Keep at it.

Ken


Top
 Profile  
Reply with quote  
PostPosted: February 7, 2020, 5:08 pm 
Online
User avatar

Joined: July 17, 2008, 9:11 am
Posts: 5343
Location: West Chicago,IL
How exactly do you determine the COG in 3 dimensions? Anything based on guesses, results in a guestimate answer. That is a lot of work IMO to get a questionable answer you could have guessed at in the 1st place.

_________________
Chuck.

“Any suspension will work if you don’t let it.” - Colin Chapman

Visit my ongoing MGB Rustoration log: over HERE

Or my Wankel powered Locost log : over HERE


Top
 Profile  
Reply with quote  
PostPosted: February 8, 2020, 12:03 pm 
Offline
The voice of reason
User avatar

Joined: January 10, 2008, 4:47 pm
Posts: 7652
Location: Massachusetts
You mention the issue of front wheel angles, that's a problem here because it changes the forces at the front and you can't model the directions with your setup.

So you can do just the front wheels straight ahead, which would be max cornering.

Once you do this you'll need some math and diagrams to figure out what is happening. Since you need that stuff you can do it ahead and save the actual effort. Plus doing it ahead means you don't build yourself into a corner.

I think it makes sense to try and model suspensions to understand the geometric forces for anti-roll, dive and squat etc. But that can be very simple stuff.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 11 posts ] 

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
POWERED_BY