LocostUSA.com

Chuck
Very perceptive.
John
I had thought about a number of ways of doing it but I really don't like doing the same thing twice or 20 times. Call me arrogant.
Tim

Ok, my bad I haven't up dated this log for far too long and I have made some design side progress. I keep a separate log which I can copy over, leaving out the bad language.

Here goes:

31 December 2011
While looking for some pics to post on the Locost site I came across the CD I bought as part of the Westlake XI build manual. The XI is similar to 7 (surprise) so a lot can be learned from these pics. One example is why the RHD XI uses floor mounted pedals and the LHD XI uses pendants. The Sprite exhaust and carbs occupy that space.
I am currently struggling with how to get going. The long single focus house project has made it hard to shift gears.
That’s about it for 2011.

3 January 2012
Pay dirt! While trolling an “old school” build on the Locost site. It was mentioned that if you use trunion type uprights you are stuck with mounting the A arms at an angle (≈5° from horizontal) in order to get the necessary caster. It would have been a pretty nasty mistake to have missed that que.

4 January 2012
I keep trying to design a Satchell link to locate the rear axle laterally. This would give a lower roll center than the usual Panhard rod. The difficulty is that the front mounts are in a place where there isn’t any structure.


12 January 2012
I made myself a target completion date of 1 June and since I did that nothing has gone well. I figured if I did this it would be good to have someone else rebuild the engine. So I have spent hours trying to find a shop to do it. Probably enough hours to be about 25% done if I had rebuilt the thing myself.
I also have been working on the rear axle layout. This is a problem as the thing is too narrow to easily mount a four trailing arm suspension arm on. I don’t want to be welding any brackets in the bearing area so I have been trying to disassemble the axle to find out where that is. Everything is corroded to dust and anything the former owner touched he torqued the hell out of. Disassembly is not going well.
To check how everything was going to fit in the 80” wheelbase I reassembled some of the front suspension and rack to check the steering wheel and seat locations. Two things were learned: 1. the seat can be quite a bit further forward than I thought which will allow those trailing arms to be mounted further inboard. A good thing. 2. Fooling around with the massive brake/clutch pedal assembly I bought from Wilwood convinced me there is no way this would fit in a 7 foot box area. So I found a stock pedal assembly on ebay. After it is cleaned up and painted and mounted on a seating buck I will be able to establish the location of quite q few things.

I love that target date
More later

10 Feb 2012
I’ve spent some time looking at the Locost site for software for designing suspensions. It is all on PCs. So this means looking for a way to run it. There is a program called “wishbones” which was written for DOS and has been converted to windows but folks seem to have quite a bit of trouble making it work. There is also an outfit called Performance Trends that sells a version for $79 that runs on XP which I have a copy I could use. I am undecided as to what to do as learning and loading the program and the data it will need is a big time sink. I have also ordered the Race Car Vehicle Dynamics Milliken book with worked out examples. I think I will wait on the software until I’ve worked thru some samples. Another big reason for the hesitation is that in looking at some Lotus 11 front end pics they seem to indicate that the A arms don’t need to have such a large variation in length as is used on most 7s. This would make the front end a lot easier to fabricate. This is because the A arms need to be mounted on a slope equal to the caster when using the Sprite trunnion type suspension mounts.

So that picture isn't your suspension? Because it looks like it may already be designed

Another "analysis" option that's a lot more fun is to make a "string computer" a la Staniforth, or scale parts out of cardboard and vary the pivot points with pushpins, and observe the wheel behavior directly by moving the parts around.

_________________
------------------------------------
Noah

Simplest is frequently best or at least more fun.

20 Feb 2012
On a trip to Iowa I stopped off in Ames and went to visit the FSAE lab at ISU. I met the suspension designer for last year’s car. He said they now have some whiz bang suspension software that can model everything including springs and shocks. He said they used to do it using 3D cad. This and a comment from a guy using an MGB front end on the Locost site have convinced me to go back to the cad method and try to more or less duplicate the Sprite set-up.

27 February 2012
After study of the Milliken books (all 3), the 2 Smith books and the Staniforth book and spending hours laying out various suspensions I think I finally have a good grasp of what it takes to make a kinematically good suspension. It reduces down to one that keeps the roll angle less than the camber change which prevents the loss of camber thrust. That is, it keeps the camber of the outside wheel from going positive in roll. There is some concern about the roll center moving around laterally but my design is constrained by using the Sprite parts so that doesn’t come up so much. The biggest problem which I haven’t yet solved is how to mount the inboard ends of the A arms so that their axis of rotation is parallel and rotated 5°-7° from horizontal to give the correct caster angle. There are similar designs out there so it should be doable.
I have just spent the last three hours didiling with the CAD method. As I am using the Sprite upright and want to maintain the Sprite track this leaves me with the top A arm length and angle as variables. My static layout looked pretty good. I assumed a 5° roll and found a reasonable roll height at 4.13”. The problem was the camber gain on the outside wheel was pretty whimpy ≈1.9°. So I started redrawing with a shorter top A at a greater angle. This gave a big jump in roll center height. That doesn’t strike me as something good to be varying a great deal as the distance from roll center to the CG is the roll torque arm. A big variation sounds unstable. This was really bugging me and then it dawned. I had not been re drawing the static roll center for the various A arms. Dah! Now my roll center varies by .1” with a 3.9° camber gain. Close enough.

28 Feb 2012
One more go around and finally get a camber gain of 5.4° with 5.4° of roll and a roll center height change of less than .1”. The only problem is this leaves the vertical distance between the inner A arm mounts at ≈4.75”. In as much as a large torsional load will be taken out between these mounts that doesn’t sound so good. Also, photos of the Lotus 11 front end seem to show the upper arm level and the lower arm in slight droop. And a good deal closer to parallel and further apart inboard. A reread of the Ludvigsen book on Chapman suggests that what is used on today’s XI may be quite different from the original. It indicates the use of parallel A arms. One should keep in mind that early Chapman layouts were in the process of first development and therefore possibly not the best. Also the tires they used and the ones I intend are pretty narrow and thus not so influenced by camber change as modern gumballs. One more whack to move those inner pivots further apart.

If anyone is interested in this please reply. I could use some feed-back on my methods and conclusions.

You said the FSAE lab used some whizz bang software, but didn't say what it was.

_________________
mobilito ergo sum
I drive therefore I am

I can explain it to you,
but I can't understand it for you.

I am not sure he told me but I suspect you could find out on their web site.

I can't really offer much in the way of feedback and/or opinion, but I have been following your thread and I do find it interesting. I do believe threads like this are useful, and have the potential to be even more valuable over time.

_________________
.. in the world

My only comment is that I wouldn't worry so much about getting 100% camber gain, since you can always start out with negative camber.

Do you expect 5 degrees of roll, that seems like a lot. I am using 2 degrees in my planning now. It would be nice to get 100% camber gain in roll, but then you don't have a compromise for braking in a straight line, single wheel bumps or accelerating. SInce your using narrow tires, the camber gain matters even less. So consider some more moderate goals and you'll end up with better mounting location and length for your upper wishbone.

_________________
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.

erioshi
Thanks for the encouragement. Keeping a log is time consuming but a really good idea when your memory is as bad as mine.
nick47
I have pretty much decided to forge ahead with what I have after reading about the early days at Lotus cutting off the suspension and trying something else on a regular basis.
Marcus
5° is a lot of roll but it gives 2" of bump travel from static and that doesn't seem like too much to me. This will be a road car and we have some pretty serious bumps here on the frozen tundra. The perfect being the enemy of the good I think trial 9 looks like it.

You should have and expect to use 2" of travel without any problem, especially if roads are bumpy.

Your may have more then 5 degrees of camber whenever you use your brakes hard though. That doesn't sound good. Something like 50% camber correction would mean you just have 2.5 degrees under heavy braking and that sounds like a better compromise. You can prevent the car from rolling 5 degrees with anti-roll bars.

_________________
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.

Roll bar schmoll bar.

Way back when I worked up a spreadsheet based on staniforth and with rather soft springs, and no roll bars, got less than 1.5 degrees roll in 1 G turns. For 5 degree roll you would have to use coat hanger wire for springs ....

I lost interest in what was rather an academic exercise given my glacial build progress.

_________________
Warren
Isuzu Pickup/SR20DE, +401 COLD frame
Build Log: viewtopic.php?f=35&t=11601