gemellocattivo.com

Which means "Evil Twin". Lets see your projects where you change boring into fun or create the fun from scratch.
It is currently Wed Mar 20, 2019 8:15 pm

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 10 posts ] 
Author Message
PostPosted: Thu Jan 24, 2019 7:52 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
Reposting this here because why not, it's interesting. It's sort of chronological look at my thoughts.

So, what I've wanted for a while now is something that is 1st gen RX-7 shaped, but without any of its weaknesses. And with more power, but still rotary - see, I'm of the view that rotaries are great for roadrace use given the ability to have a nice long TBO as long as coolant and oil are kept in the right temperatures and obviously supplied.

So I came up with this.

Image

That was a few iterations ago but it's already nice and uploaded. There's also a couple (very) crude overlays to show how everything fits - yes, they're even a bit further back in the process than that last one but they just give an idea.

Image
Image

So yeah, the basic idea is an SCCA GT-2 car running a 3-rotor, chassis design similar to current Trans Am practice. But similar to some other cars that people have seen I was thinking in terms of keeping consumables affordable - Ford 9" rear, Late Model brakes, Jerico (or something) trans, etc. And now the crux of the matter...

Image

See, nice affordable Wide 5 hubs. Real Racing Wheels has 16" aluminum wheels in the right sizes for far more affordable than BBS centerlock stuff, everything is golden.

Except.

Image

10 degree KPI to get the scrub anywhere near right and it's still not great. Ugh.

Wasn't sure whether to just put up with it, or start designing my own centerlock uprights so I disappeared down a couple different rabbitholes. First though I taught myself surfacing to get a better bodywork setup:

Image

Image

Image


Last edited by mekilljoydammit on Thu Jan 24, 2019 8:03 pm, edited 1 time in total.

Top
 Profile Send private message  
 
PostPosted: Thu Jan 24, 2019 7:54 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
So then I found some centerlock parts and reverse engineered them:

New (to me) parts time! I want to thank the guys at Roadraceparts.com for working with me and selling me these used bits so I can measure them up; it's an incredible help. So that said, on to pics!

Image

So that's a BBS E55 series wheel center setup for centerlock, and a used centerlock spindle from I'm not sure what honestly.

Image

Beefy sucker; the bearing races on the left are something like 2" but the whole thing is hollowed out pretty extensively. It's not even just a straight bore either, the part that bulges out to locate the wheel is a larger ID than the part that the bearing races are on. I need to put some thought into how they did that.

Image

Side view of the wheel center... one of the older style where the rim halves bolt on to the outside of the wheel center. The newer ones are setup for bunch more caliper clearance but this lets me confirm all of the machining surfaces necessary to bolt it to the spindle.

Image

Rear (or inside, whatever) view of the wheel center; note 12 holes for the drive pins. Same wheel center works with spindles with either 4 or 6 drive pins.

Image

CAD model has it as within an ounce or so of the real one, so nice confirmation I probably got the internal geometry right. Already have my redesign to take angular contact bearings done - this weekend's project is to hash out the suspension uprights and brake package.


Top
 Profile Send private message  
 
PostPosted: Thu Jan 24, 2019 7:57 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
Of course, centerlock is too expensive (cool but good lord, the cost of whittling a piece of steel to shape) and isn't necessary especially for sprint races where I'm not really doing tire changes... ended up talking to Pancho Weaver at Road America's Trans Am weekend and saw how he did the Challenger Boris Said drove.

So had the thought to set up a 5x5 setup with a BBS styled 3-piece wheel to see if the geometry could get there.

Old design front wheel assumption was Wide 5 - these are cheap and cheerful.

Image

The problem is they're not intended for low scrub applications - with the biggest backspacing wheels available the rotor is inboard of centerline.

Image

I don't like that at all - it forces more scrub, more kingpin angle, and mostly, both. See, where the lower ball joint (or rod end, whatever) goes is determined by the inner face of the brake rotor. So I went and mocked up... well, I started with the idea of a BBS E50 style center but more concave, like so:

Image

Which not only looks the business but helps move things outboard a lot.

Image

Brake rotor is moved out a full 1.6" plus it looks snazzy! I don't think it will hold me back any compared to centerlock, really. And I think I could even just *gasp* buy some E50 centers and use them as is if I were to win the lotto or something... whatever, it may be cheaper to just clone them out of billet or something. The point is it should work. Also, new setup is with bigger rotors and bigger calipers which is also good. Designing around 13.06x1.25" rotors and Wilwood GN6R calipers to start, maybe better calipers later. Should be plenty of brake for a car with a 2300 pound minimum weight right? OK the horsepower target is 450 at the wheels so it's not like it'd be slow.

Then did some rejiggering of the chassis. Converted everything to round tube, threw a dummy in to make sure of positioning... well.

Image
Image

Left the rear chassis tubes off the redesign for the moment since I need to actually put steel in the right places. Chassis lost 130 pounds compared to the last version and should be a lot stiffer


Top
 Profile Send private message  
 
PostPosted: Thu Jan 24, 2019 7:58 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
So then, got a lifed out 5x5 hub in; probably Howe or something. Reverse engineered it to get it for layout in Solidworks.

Image

Quick whipped that up. It's hilarious how big the bearings are compared to production car stuff... the taper bearings are 2" OD, 3.5" OD. It positively dwarfs any of the stock Mazda stuff.

Next, decided to do some FEA on the chassis weldment.

Image

So, what's going on here? The four joints at the front bulkhead were set to immovable and a 2,000 pound load is applied to the rear coilover mounts, one upwards and one downwards. Some math gives a result of just shy of 20,000 ft-lb per degree of chassis twist, which is a bit more than the Viper GTS-R (the caged race ones) and in the same ballpark as a Zonda, Ford GT, or a modern F-1 car. For a non-aero car I figure it's doing pretty OK.

OK, why bother? For one thing it's a good project to get my feet wet with FEA. :D More seriously though, I want to try to cut as much weight out so this means I've got a reasonable tool (a lot better than guesswork) to figure out how to do so.


Top
 Profile Send private message  
 
PostPosted: Thu Jan 24, 2019 8:02 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
Where I'm at now... I realized that doing a first cut for GT-2 would be awesome... but expensive. Competitive power in GT-2 would take a 3-rotor, ideally a 6-speed sequential trans, big tires big wheels and either custom wheels or CNCing centers.

And you know, maaaaaaybe I should start at a lower budget class. You know, like one where I have the parts for 3 shortblocks that should be relatively competitive sitting around. And for all the obsessing about scrub radius it's faster to get on track with Wide 5, where I can get nice light affordable wheels. So I started paring down the chassis.

Image

Long story short, it's about half the weight and 2/3rds the chassis stiffness. Next step is suspension geometry.


Top
 Profile Send private message  
 
PostPosted: Fri Jan 25, 2019 7:59 am 
Offline

Joined: Thu Jan 01, 2015 6:47 pm
Posts: 2519
mekilljoydammit wrote:
Next, decided to do some FEA on the chassis weldment.

Image

So, what's going on here? The four joints at the front bulkhead were set to immovable and a 2,000 pound load is applied to the rear coilover mounts, one upwards and one downwards. Some math gives a result of just shy of 20,000 ft-lb per degree of chassis twist, which is a bit more than the Viper GTS-R (the caged race ones) and in the same ballpark as a Zonda, Ford GT, or a modern F-1 car. For a non-aero car I figure it's doing pretty OK.

OK, why bother? For one thing it's a good project to get my feet wet with FEA. :D More seriously though, I want to try to cut as much weight out so this means I've got a reasonable tool (a lot better than guesswork) to figure out how to do so.


Love it. This was my what I did my design project on in school...chassis design and FEA analysis of it so comments, not criticisms.

You can't use any immovable constrains. You are shooting for a TOTAL of 6 constraints, immovable is 6 just at the 1 point, so with 4 points thats 24 degrees of freedom contained. That means the constrains them selves and making a significant contribution to the design strength and stiffness, and the real frame will probably flex significantly more the the analysis is predicting. I can't find any pics but what I settled on for torsion testing was the moment constraints anywhere, x/y/z at 1 corner, y/z at the next, (these where the 2 rears iirc), z at a 3rd and itr's not fully constrained. then apply a z load at the 4th and you have a pure and clean torsion measurement. Using this method the final chassis measured within 2% of the analysis result.

The other thing that pops out is very non-uniform loading. The work I did focused on chassis stiffness/lb of chassis weight and the basic result was that the more uniform the loads end up, them lighter the chassis can be to obtain the same overall stiffness. In the pic you posted that would mean thicker tubes in the front and back and thinner (I used 0.028" because it was the thinnest I could get.) everywhere else.

I also see a lot of un triangulated spaces like the very front opening. Ideally EVERY point is supported in 3 directions. Have a look at the chassis below, I play around with removing just 1 triangulating tube from an external surface and frame lost 3/4 of its stiffness. You have a lot of redundant members so your result wouldn't be that drastic, but it makes a big difference and lets you go smaller and thinner on the tubing.

But often the rules and min weight limits make caring about these tiny details a waste of time so there you go. :)


Attachments:
20180203_205706.jpg
20180203_205706.jpg [ 185.23 KiB | Viewed 269 times ]
Top
 Profile Send private message  
 
PostPosted: Fri Jan 25, 2019 10:06 am 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
The constraints on the FEA were done by dint of "hm, this should work" so I appreciate the advice on how to do better. I had FEA courses in school back in the day of course, but it's been a while since then. I'll try redoing that at some point in the near future. My thinking was that I'd constrain the front bulkhead and apply Z loads to the rear shock mount points - I'd of course prefer "more accurate" but I do feel that it let me figure some stuff out with what tubes are important and not.

The last frame is more optimum; the darker grey tubes are required safety structure at 1.5x0.095" wall tube, the lighter grey tubes are 1.5x0.065". I didn't want to go much lighter than that as part of my goals is to have something that can survive contact without too much repair and formula car style wall thicknesses get pretty dent sensitive. There's some stuff that isn't obvious just looking at the tubes either; the front is closed by the "engine mount", which in typical fashion for these sorts of cars is a piece of billet aluminum carrying all the accessories. Doing it that way vs triangulating with tubes aids serviceability a lot too, as you can just pull the whole driveline out the front. I didn't really want to triangulate through the windows for obvious reasons, and a lot of other areas removing tubes didn't seem to have much of a stiffness influence; the floor for example.

Given minimum weight (2080lbs with driver) I'm leery of adding too much more in the way of tubing, especially as I'm also concerned with serviceability. I still have a few more things I want to try but even as is I frankly think I'm doing a lot better than most of the existing cars in the class.


Top
 Profile Send private message  
 
PostPosted: Mon Jan 28, 2019 7:34 am 
Offline

Joined: Tue Jan 06, 2015 8:09 am
Posts: 107
Okay, I am following, this looks fun.

_________________
Wade Williams
Master Ferrari Tech
Restoration Tech


Top
 Profile Send private message  
 
PostPosted: Mon Jan 28, 2019 10:35 am 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
Thanks; I think I saw in your thread that you're crew chief for a TA2 car? The original GT2 version is classed the same as those in SCCA so similar performance envelope - the GT3 version is a lot less power so I'm lightweighting stuff. Once stuff gets out of digital and into physical, you may recognize a lot of parts. ;)

My overriding concept is that I'm tired of dealing with unibody race cars, so I want something where hitting something means I pop new bodywork out of my molds, weld some tubes, and replace some relatively cheap commodity circle track suspension parts.


Top
 Profile Send private message  
 
PostPosted: Tue Jan 29, 2019 12:57 pm 
Offline

Joined: Wed Jan 23, 2019 11:13 am
Posts: 24
Oh yeah, I suppose I should mention one of the more ambitious parts of what I have in mind.

So I'm a test engineer as my day job; suffice to say that I'm comfortable with large amounts of data and processing it. I got started racing in my 30s so I'm never going to catch up in experience with guys who started as kids in carts... but what I can do is throw disproportionate amounts of engineering at things. :lol:

I'm of a mind to do programming and hardware design on the DAQ myself as a learning experience and honestly because it's not that hard. In addition to all the health monitoring stuff (all the driveline temperatures and pressures and so on) I want at minimum on each corner tire surface temperature, wheel speed, and shock position. In an ideal world also tire pressure and brake temperature but I don't know a good inexpensive way to do tire pressure. Add to this steering angle, brake pressures, ride height, plus the normal accels/gyros/GPS and in an ideal world (if I can figure out a cheap way of doing it) optical flow sensors at front and rear. Driveline torque sensors would be nice, along with some other stuff, but again starts getting into the realm of expensive sensors. Needless to say this takes me into the realm of either quite expensive DAQ options, or daisy chaining a bunch of inexpensive stuff together, or both. Thus more impetus to roll my own.

What's the point? Well, I'm rolling my own chassis already so I know the suspension points... if you think about it, from that pile of sensors I listed, you can calculate what the forces seen at the tire are, vertically and horizontally, but also come up with a pretty good estimate of load on the tire, orientation relative to the track, and slip angle. All of which is to say, you can start building an accurate tire model! And if you have an accurate tire model and know all the physical parameters of the rest of the car (more or less anyway) you can feed it back into a vehicle dynamics simulation, validate the whole thing... and start changing parameters.

The idea isn't to assume "the software will spit out the optimum setup" but come up with a tool to sorta... narrow down the possible changes to a few promising ones that are testable on track days or whatever. After all, if you can move the suspension mount points without too much trouble, the list of possible changes gets pretty big pretty fast.


Top
 Profile Send private message  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 10 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

Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group