Suspension program - Vsusp

[Rob7]

I've used the trial version of Suspension Analyzer and in affect I am comparing Vsusp to that software.

Pro:
Vsusp looks like it will eventually be in that class of software. It will be very interesting to see how this software progresses.
The GUI, dimensioning displays and chart are very good.

Con:
When you make changes to the chart X data, the chart doesn't change unless you change one of the things you want graphed. I expected an "Enter" or "Redraw" button to appear after a change so the changes would be applied without changing a Y axis item. By having a button, if you don't click on the button the changes will be ignored and the original values would be used when you do something else.

There doesn't seem to be any way to input steering changes with resulting toe changes due to bump/droop/roll [yet?]. Maybe have a top view to display toe changes in 2d until the 3d version is done?

The testing you guys are doing will help me fix the problems more quickly. I put the chart redraw problem on the list and will get to that before too long.

Steering/toe: To make the program show steering properly, I'd have to add a third dimension to the steering knuckle and also have the steering axis in 3D. It seemed like too much bad hacking to try to do that in 2D. The closest I could get would probably be to show how the tie rod end moves laterally in comparison to the steering axis during bump/roll .. hmm

[horizenjob]

Here's a start on Vsusp setup for Car9. Car9 is a project to produce an updated Locost style chassis suitable for track use. My thought is that a track car works better on the street then the other way around. I am trying to collect and produce enough 3D models to make other home builder's projects easier.

I think we still have a problem with the accuracy or resolution of the data used in Vsusp. It's probably two problems, but they are perhaps related. When the data is exported and imported it is rounded to the nearest millimeter. That is 1/25" or .04", but a common 1/2" rod end has 20 threads per inch. You can turn them 1/2 turn so they produce adjustments of 1/40" or .025". Many suspension offer much finer adjustments with left and right hand thread joints or other means.

In the file below when I saved the link the numbers where changed, so when you load it and look at the front suspension it has different camber and angles on the control arms. For instance the lower control arm was level, then after I saved it it now has an angle of 0.3 degrees.
http://tinyurl.com/Car9-G-V-001

The way I am trying to use this is to have Vsusp in multiple tabs. Then I make some changes and position things so the graphs overlay and I can snap back and forth between tabs and see how the graphs compare. So I would like to have exactly 0 camber at static, but the best I can do now is get within 1/2 degree.

For the data entry boxes I think you need to tolerate 3 decimal places. People can easily measure things to within 1/8" or 0.125". I can understand an engineer thinking measurements should be in 10'ths and 100'ths, but not every home builder will be comfortable with this. I don't know what limitations the software you use to handle those number boxes has though. For me the ability to click up and down thru the numbers is less important then being able to type in an offset in the description of a spindle and have it stay the same next time I load from the link.

Thanks for being willing to work on this. I will be happy to use links to this from my threads and website on Car9. I think it will be a great tool for our members to use on their locost projects.

[Rob7]

All sounds great. I haven't yet fixed the 1mm bug, but I'll get to that in the next couple days. At the moment I'm fixing another bug which is causing problems for internet explorer, but that should be done by tomorrow.

I can also add some precision for the inputs. I don't see any problems with increasing it to three decimal places. Thanks for your patience in the meantime

"a track car works better on the street then the other way around" That interesting, could you explain a little more on that? Or actually, I can probably read through your threads a little later today

That also reminds of something I read a while ago, I think it was in one of Carroll Smith's books - something like "your track car should be easier to drive than your daily driver"

[olrowdy_01]

"a track car works better on the street then the other way around" That interesting, could you explain a little more on that?

With a full race cam in a stop/go traffic jam?

That also reminds of something I read a while ago, I think it was in one of Carroll Smith's books - something like "your track car should be easier to drive than your daily driver"

Somehow I can't visualize trying to keep up with 200 mph rush hour traffic on the interstate while driving a Smart Car.

On second thought, I maybe I can.

http://www.youtube.com/watch?v=g5Yo6V2E0sM

[andyawesome]

Thanks alot for all your work Rob, and it seems like serendipity to me time to build! and thank you cheapracer for messaging rob, Doh, why didn't I think of that. Anywho I think the program is really good and is perfect for what I was looking for! Now for more playing around.

p.s. I don't know if its just me but I can't enter numbers from the keypad no big deal though

[horizenjob]

I can also add some precision for the inputs. I don't see any problems with increasing it to three decimal places. Thanks for your patience in the meantime

I'm grateful your putting in the effort. This will be a good help for our builders, we can share our ideas more easily. I think many builders land up putting a lot of hours into this and I hope this will help us publish some standard setups. It's not a full package, but it's a big step up from drawing or string computers.

My thought is that a track car works better on the street then the other way around.

You always get in trouble when you generalize and you should never put a luggage rack over the fan on top of a Porsche 917. In the context of the cars we build here it is a sensible statement. If you look at Car9 and want to get a sports car for street use, say picnics, weekend rides to car meets or average daily use - you can make a couple simple changes like raise it 2" and skip the roll bar braces going by the driver's head. It wouldn't have any downsides compared to a Seven or Locost or your average motorcycle.

Since the Locost was not designed for the track to start, it has a less robust frame with less protection for the driver. That's not so easy to fix once you built the car and want to track it. The builder is somewhat more on their own when it comes to suspension , especially IRS. Car9 is designed for IRS, but the design is just as suitable for a live rear axle because it does not depend on a suspension box in the rear.

To best support the builders and drivers I am trying to provide a very adjustable suspension for Car9 with data on different setups. The wishbones are attached with bolt on tear tabs. Replacing the tabs with a different shape or just flipping them upside down can adjust roll centers, adjust anti-dive and anti-squat / anti-lift and help with ride height etc. When we have data and people know what they want, builders can just weld the tabs on for street cars.

Vsusp looks like a good way to explore this stuff and to demo ideas or dimensions for setups in threads here on LocostUSA and on the Car9 website. People will be able to click on links and see suspensions with parts from Miatas, Subarus, Fords and contribute their own ideas.

An example of something I'd like to explore and make easier to look at is the effect of drop pins on spindles. Production car spindles, front and rear, seem to suffer from having the lower wishbone too high off the ground for us. I think that can effect the choices or tradeoffs between Roll Center heights and camber control. I don't really know yet though. How much better will the numbers look with a drop spindle?

We can design some simple spindles that use bolt on hubs, how much better would they be? Vsusp can be a quick way to look at some of this stuff. It turns out Car9 is especially easy to do here because my wishbones have one leg perpendicular to the chassis, no extra math involved to work in 2D.

[cheapracer]

Just a personal wish item, would like to be able to 'grab' a button that I could dynamically roll and bump together as one with my mouse.

it took a guy in China

No it was me, Sum Yung Guy operates our lathe, he can't even use a computer.

[Rob7]

Just wanted to do a quick check in to let you all know what's up.
I'm probably halfway through some uninteresting but necessary fixes for vsusp which involve script loading and namespacing. After this, vsusp will work again in internet explorer.
Today and tomorrow, I have to do some extra hours for my day job. But progress on vsusp will speed up again after. I'll be replying to all the messages and addressing all of the issues you guys have found.

Cheers,
Rob

[mattrogers]

I'm probably halfway through some uninteresting but necessary fixes for vsusp which involve script loading and namespacing. After this, vsusp will work again in internet explorer.

Interestingly, Kogan, an Australian retailer, is now imposing a surcharge for orders placed through Internet Explorer 7. Because, as you've found out, it takes more effort to program things correctly for IE.

http://www.zdnet.com/blog/networking/au ... 7-tax/2496

[olrowdy_01]

........... Interestingly, Kogan, an Australian retailer, is now imposing a surcharge for orders placed through Internet Explorer 7. Because, as you've found out, it takes more effort to program things correctly for IE.

http://www.zdnet.com/blog/networking/au ... 7-tax/2496

Wait till Windows 8 and IE10 gets turned loose in the wild!

"Look maw, I've got a 30 lb [desktop] tablet with a 21" monitor!" And it has an on screen keyboard built in too.

[Pe7e]

Do folks mind if I create a thread for suspension discussion using Vsusp and move this discussion there? Then we can discuss Vsusp here and suspension design using Vsusp in the new thread. That should help Rob keep track of our feedback better.

+1 on the Designing with Vsusp thread, if Rob agrees.

Pete

Mod edit --- go here for actual suspension design as opposed to Vsusp software discussion: http://locostusa.com/forums/viewtopic.php?f=5&t=13952

[Rob7]

Do folks mind if I create a thread for suspension discussion using Vsusp and move this discussion there? Then we can discuss Vsusp here and suspension design using Vsusp in the new thread. That should help Rob keep track of our feedback better.

+1 on the Designing with Vsusp thread, if Rob agrees.

Pete

Sounds great to me if you guys want to do that

[Rob7]

A new version of VSusp is up:

-fixed initial bump units (thanks Olrowdy_01, Horizenjob)
-fixed chart redraw on changed settings (thanks Olrowdy_01)
-fixed draw toggle redraw (prefs panel)
-boring technical fixes (namespace and script loading - was breaking IE support)
-slightly larger control panel tab font

Coming soon:

- will fix the 1mm roundoff problem
- will replace the disclaimer window with a less intrusive Terms link

[Rob7]

Thanks alot for all your work Rob, and it seems like serendipity to me time to build! and thank you cheapracer for messaging rob, Doh, why didn't I think of that. Anywho I think the program is really good and is perfect for what I was looking for! Now for more playing around.

p.s. I don't know if its just me but I can't enter numbers from the keypad no big deal though

You're very welcome. The interest so far in Vsusp is exciting. I'm also getting ready for my own Locost build - I ordered a Millermatic 140 and other goodies yesterday.

About the number pad - yes, that's a big downside to the input box 'spinner' I used (which enables arrow keys and the up/down mouse clicks). But I may be able to get it the numpad working and still keep the input boxes tweakable with the arrow keys. I'm adding that to the TODO list

[Rob7]

I like how the RC moves inside on turns - reduces jacking.

I am not completely comfortable really using the roll center numbers because I am not sure what they really are saying. In particular I am dis regarding the horizontal location information. I do look at the angle of the line of force from the contact patch to the instant center. I think that's a direct indication of the jacking forces. The roll center stuff produces results that are not sensible in some cases, so the fact that they are sometimes sensible does not really make up for this...

Actually I think it would be nice if Vsusp could plot that angle. Either directly or as a percent of cornering force showing up as jacking. Does that make sense?

The Roll center numbers should be distances x,y from the ground/centerline. Maybe the origin should be the roll center at zero roll / zero bump? Or are you seeing very strange coords in your project?