LocostUSA.com

Go on! You can do this stuff in your sleep, Perry. With the speed in which you finish a project, I sometimes wonder if you DO work in your sleep.

I'm just stumblin' and fumblin' my way through things, but getting to a better understanding of what I need to do. In the end, I seem to do OK.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

Indeed, Chuck. That first setup would have run out of vertical motion real quick.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

I spent the afternoon sitting at my desk (actually my portable welding table cover with cardboard top) right at the front of the build table. This is the way I like to work. I sit there, try to visualize what I think would work best, mock it up somehow using wood, plastic, cardboard, or whatever works. Then evaluate how well I think it will work, and can I actually make what's required?

I have a true one-off build, and I've got to solve some unique issues with my donor parts. I have some really cool software to help with the designing, but you've got to feed it real data to get meaningful results. That means mocking things up, measuring from certain datum lines I will set up, and inputting the accurate data to the software. So, I'm going to embark on a little development program, which will take me some time. Given all the variables, and the many iterations I'll need to try, I'm thinking it's going to be a month (at least) before I have a finished product in metal.

That's OK. I gave myself permission to take that time this afternoon. Otherwise, I'll push too hard to get something quick, and make too many compromises. I don't want to do that. I've got years into this project. With COVID-19 isolation, I don't have a lot of other stuff to do anyway. I'd rather be thorough and get a good result.

I like to visualize/mockup/measure/observe as a methodology. I find I arrive at better solutions that way. One little issue I rediscovered today is the steering rack location and tie rod angles. With the wheelbase I decided on, the rack has to be just on one side or the other of the front frame. The realities of that is that the tie rods will need to sweep slightly forward, or slightly backward, depending on which side of the front frame I mount the rack to. The long rule in the photo simulates the general arrangement if the rack is at the back of the front frame - almost certainly where I'll put it.


It's going to be a challenge to get everything lined up as needed to avoid all steering effects. I've always planned on a custom rack, and that's probably the best solution here anyway. So, tonight I'll make myself a nice drink, and tomorrow, start my development program.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

A little update. I'm using Perry's method of setting up for the front suspension with a slight modification. I dropped the tire and wheel platform to simulate ride height rather than raise up the chassis to ride height off the build table. All else is pretty similar.

I created a vertical piece to hold the spindles. When joined with the flat base, the center of the spindle is at the correct height.


The spindle/disk/balljoint combo is pretty heavy. So, I created the remainder of the jig to both place it correctly and not break or flex.


This side faces out from the car centerline.


This is the side of the jig that faces towards the chassis centerline.


Here the upper and lower balljoints are removed from their screw-in housings. I plan to use them in this fashion to compute the various angles required to design the suspension. I'm going to use 3 methods, and see how close we get with each one. I don't have any real precise measuring devices, but I think I can get to half a degree at least.


I'm going to use a similar technique to calculate the SAI, scrub radius, etc from this view.


Happy Thanksgiving to all in the USA (and Yanks abroad too).

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

Looks good Lonnie, thats exactly what I do without mathematical wizardry and computations. I use a digital level and cardboard mock ups. The end result is the same, that's what counts.

_________________
Perry

'If man built it, man can fix it'
"No one ever told me I couldn't do it."
"If you can't build it safe, don't build it."

Perry's Locost Super Che7enette Build
Perry's TBird Based 5.0L Super 7 L.S.O
Perry's S10 Super 7 The 3rd
Perry's 4th Build The Topolino 500 (Little Mouse) Altered
Perry's 5th Build the Super Slant 6 Super 7
Perry's Final Build the 1929 Mercedes Gazelle

Here's my latest experiment with measuring the spindle dimensions and ball joint locations. As may be obvious, there is no way to directly measure the distances or angles between some spindle/balljoint features with typical home shop measuring tools. Unfortunately, those measurements are required for suspension design purposes. If I owned a good height gauge (18" perhaps) and some very large dividers or calipers, it would be possible to measure or calculate in most cases. But, I don't, so I'm doing some indirect measurements.

I was able to put my small digital caliper on the ball joints and measure the diameter of each at their equator. Using some magnifiers, that location is actually pretty easy to see from machining marks, and then mark with a black Sharpie. I decided to see if I could use 2D photogrammetry for some, if not all, of the measurements.

So, I created some simple targets, and simple scales in a drawing program.



I'm still doing some experimenting with my setup. I'll upload some photos of that probably tomorrow. It was too late to get real clear photos because of darkness, but the ones below Illustrate the basic ideas.




I've located a great photogrammetry software that has a trial period. It even corrects an image for based on the focal length of the camera lens to make a better 2D orthographic. I'm not sure that will be necessary. I'm going to try them in the software I have now, and the attempt to verify the measurements using indirect methods in real life. It's going to be more approximate than I'd like, but if I can get to 1/16" (not quite 2mm) I may just call it good and get on with things.

This is one of the downsides to doing a one-off. You can't just build to a plan, you've go to figure out a lot of things, and then create the plan, and then build the parts.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

The first shot at measuring did yield some important data, which I'll attempt to verify today. It also revealed some minor adjustments I need to make, which I'll do today too.

It looks like the built-in castor angle is ~3 degrees, since I have the flat of the steering arm level. I was surprised to see a SAI (KPI) of 12 degrees. I was expecting about 9. I'll see if I can verify the latter angle out in the online literature for the Cobra world. That's where it comes from.

I think this process is going to work well, if I can verify the accuracy on the parts or online.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

Are you getting those angle values off the physical mock up, or off the photos? Reason I ask is because both photos both appear to not be square to the angles shown, so I suspect that there'll be a cosine error overlaid onto the values.

_________________
Midlana book: Build this mid-engine Locost!, http://midlana.com/stuff/book/
Kimini book: Designing mid-engine cars using FWD drivetrains
Both available from https://www.lulu.com/

I'm not sure exactly what you're thinking isn't square, Kurt. Everything should be square, parallel, yada-yada as far as I can make it. The targets are all in 3-space. This is really a pseudo projection of them onto 2-space. I've been careful, but these photos are without correction for the lens,which will distort a little bit. I'd have to use that photogrammetric software to do that correction, and I'm not sure it's needed yet.

I improved the target accuracy (placement) today, placed a couple of new ones, and took some new test photos. I'll be able to do some measuring comparing some of the spindle assembly to measurements of the photo projections now. If those are good, I'll be able to trust the projection measurements in software I can't get directly from the spindle. That's the plan, anyway.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

Isn't what you are calling Static Caster, really trail.

You are measuring the offset from the steering axis to the center of the spindle.

It's called trail because the spindle trails the steering axis.

Given that I'd say that is a left side spindle for a rear steer application.


I come from the cycle world and trail is used on most, some are even adjustable.
Especially choppers that have increased rake will often have way too much trail.

I'm not sure how it plays out on a car.
But I think On a car this amplifies the affects of weight jacking mostly.

_________________
Honey anyone?

Caster angle in a car is like rake on a motorcycle: it is the angle between the steering axis and vertical when looked at from the side. Trail is the distance between where the steering axis would intersect the ground and the centre of the contact patch of the tire. It is affected by caster/rake, longitudinal steering axis offset (fork offset + triple clamp offset on bikes), and the tire diameter.
See: http://www.super7thheaven.co.uk/glossar ... is-offset/
If that were a rear steer spindle then it would have a negative caster angle, which would be bad for the same reason as having a negative rake on a bike.

Lonnie, I like your technique for measuring the angles. I would recommend you try to take the photos from as far away as possible with a high optical zoom in order to reduce the parallax and foreshortening effect, which should help in accuracy.

Cheers,
Jacob

@Bent Wrench
Thanks for the input. It is actually a front steer setup. Here is that same spindle on the donor with the steering tie rod and anti-roll bar visible.



@Posthumane
Thanks Jacob.

I did learn how to improve on the calibration (setting distances in the image underlay for the CAD program) and improved the accuracy significantly. I also relocated the targets a little to make them more accurate.

I was very gratified to find actual measurements for the items closest to the center of the photo was about 1/32", which got me enthused. That's good enough for me. However, the farther you move from the center, the less accurate things become. That makes sense when you think about a lens and optics. But, I think it's too much error to trust for my suspension design process. It was almost 1/2" for the distance between the upper and lower balljoint centers. It looks like that software which corrects for the lens optics will be a requirement. I'm going to try a more pragmatic, old school approach today.

I'll see if I can improve the accuracy of the photogrammetry process too, but I'm getting into a high-effort situation, and I'm wondering if it's worth it. I may be able to do some measuring to critical locations on the spindle set in it's correct position, and let the suspension software do some of the work for me. Stay tuned.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886

Lonnie, as I re-read your posts about the front, I started thinking. So, I had to get a beer to stop the pain from thinking. . . Anyway, if you are not married to the idea of using the strut based SN95 front spindles, you could use Lexus spindles instead. I originally tried to use them but the big/bulky circle track ball joints interfered with the rotors. Thinking back, I wish I would have abandoned the ball joints and kept the compact spindles with much better brakes (2 piston floating calipers). They are the same bolt pattern and are dual wishbone. My donor was an SC400. Just a thought.

_________________
Gavin

My build: http://www.locostusa.com/forums/viewtopic.php?f=35&t=16005

Lonnie, you are probably well beyond this point by now, but Wilwood’s Mustang II spindles have an engineering drawing available on-line with all the necessary dimensions and angles. I built a spread sheet taking the Wilwood data plus some geometry calcs to generate all the input needed for my suspension design using Wishbone for 3D analysis.

But it’s a cool approach you are using, so carry on.

What photogrammetry software are you using?

_________________
Cheers, Tom

My Car9 build: viewtopic.php?f=35&t=14613
"It's the construction of the car-the sheer lunacy and joy of making diverse parts come together and work as one-that counts."
Ultima Spyder, Northstar 4.0, Porsche G50/52

Thanks for the idea, Gavin. However, I'm so far into the build, and I have all the parts for the SN95 donor (some of which were pretty expensive), that I'm pretty committed at this point. If there was engineering documentation for the Cobra conversion, I'd be set.

I'll find a way, but it's a big PITA right now due to a no good means of getting accurate dimensions.

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886