|
kastein posted:When I needed to CMM something I ended up putting it on an old shipping crate in one of the laser cutters about to be shipped at work and used the machine vision system to locate the center of each feature. Got me within about a tenth of a thou on each one. You're also helped by the fact that the designers likely worked in even increments of millimeters, or half-mm's, so if you come up with 88.4683286, you can be reasonably confident its 88.5.
|
#
?
Jan 23, 2020 18:42
|
|
|
|
| # ? May 1, 2024 12:56 |
|
|
Yeah, or for example you can say "well it's a hair under 88.5 and something presses into it so that thing must be exactly 88.5 and I need 88.5 minus pressfit".
|
|
#
?
Jan 23, 2020 21:11
|
|
|
Proper reverse engineering still needs engineering, yeah.
|
|
#
?
Jan 23, 2020 21:34
|
|
|
Aeka 2.0 posted:Any reason to sticking with the FD trigger? I guess if they work they work. I'm surprised I don't already know the answer to this, but will an FC CAS fit into an FB housing? For some reason, I seem to recall that they do. IS the FD trigger better than the FD, or just more available? I'm still unsure of where exactly I'm going with my SA22C, but it will be a 13B EFI. Which generation depends on which keg I can make build compression, which will inform which system I use. Also, I guess I need to decide on whether I'm sticking with the stock crossmember of going to the FC front suspension, as that will somewhat important in engine block and front cover choice.
|
|
#
?
Jan 23, 2020 22:27
|
|
|
FC CAS fits, is a valid solution. I just would rather not use it because they're long out of production and a kludgey looking solution to me. Probably a 36-1 wheel is more accurate too but I don't pretend that I think that's a factor for NA stuff.
|
|
#
?
Jan 23, 2020 23:04
|
|
|
Fair point. I like your idea of trimming out the fat on the front cover to begin with. Heck, build in a mount for a sensor for use with a crank trigger wheel, along with all the other mounts and ports you need.
|
|
#
?
Jan 24, 2020 00:21
|
|
|
Any reason you can't just make a piece that bolts onto the harmonic balancer and a bracket to mount whatever sensor works with your ECU over it at the right airgap? Should be easy to cad and laser cut if you know the angles for the start and stop of each tooth.
|
|
#
?
Jan 24, 2020 01:39
|
|
|
Well, no, there's no technical reason. But it's like... my goal isn't really just to have a crank trigger for a 13B with the engine mounts I like, it's actually about building a tool to reverse engineer things and work on sand casting from 3d printed patterns. Making a new front cover is just kind of a concrete reason to do it that makes me follow through, but also is simple enough that I don't have to worry about a lot of things; it's pretty much a "if it bolts up after finish machining it's right" sort of thing.
|
|
#
?
Jan 24, 2020 13:45
|
|
|
Working on various stuff... got a rotary broach to make odd shaped holes so I can make a square hole in a shaft to make a stepper motor actuator for some surplus valves I bought. Which sounds silly once I say it, but the commercial version of what I'm bashing together out of custom stuff and a couple ebay purchases is, even after tool cost, a tenth the cost of a commercial dyno automatic flow control valve. More and more looking at all the stuff to build an EProd RX-7 and just going... ugh. Like, look at all the stuff I'm working on, a tube car gets me more excited. I may just be back to the idea of building a lower class tube car first and then redoing it for 3-rotor power and more tire later. Kinda torn. EProd definitely is a class with good participation but ugh unibodies.
|
|
#
?
Feb 24, 2020 00:13
|
|
|
Another quick update... decided to hell with dicking with unibody stuff. It'd be a good class to race in but god, I just don't want to build a unibody car again. Did some mocking up, GT-3 (15x7 wheels - and about 10" wide by 23" diameter tires) fit with my GT-2 chassis, I have most of the driveline for it, screw it, let's do that. Funny, tires don't stick out that far with the stock body. Only about a 2" flare per side... easily doable with the EProd bodywork molds. Which I could just apply to one of the wrecked race car shells in the field to pull molds from. Easypeasy.  Got the suspension geometry mostly figured out... need to run some numbers and maybe fiddle stuff a bit but it looks pretty close. Scrub isn't where I'd really like in GT-3 form but oh well, the wide-5 stuff has too many advantages to ignore. Fun fact, if you take the revised chassis and take off the roll cage stuff and door bars, it's not far off Lotus 7 width. There's still design work to do, but honestly I'm happy enough with the chassis (Mk6 now) to, once the ground thaws and I can get the trailer out of the field, weld up the I-beams I got from a coworker into a chassis table and start buying round tube. For rear end, I'm thinking _eventually_ on going with what Winters calls a non-QC. So pinion offset on differentials affects power loss... if you look at Ford 9" stuff it's hella durable, but the pinion is basically acting like a worm gear. Going to stuff with lower offset like the Ford 8.8 or GM 12 bolt has been dyno tested to be worth about 3% power at the rear wheels, but both of those axles don't have dropout carriers so they're heavy and a pain in the rear end. Quick change differentials use a bevel gear, which is even more efficient but the change gears hurt. Well, Winters makes a setup that's the QC ring and pinion but turned 180 degrees which should be most efficient and is off the shelf. I figure every little bit of efficiency helps. Short term though, the cool thing is I can make everything based on whatever I can find used or cheap... revalved Bilsteins, used 9" housings and center sections, Wilwood brakes... and upgrade piecewise to the "better" stuff.
|
|
#
?
Mar 14, 2020 17:46
|
|
|
Hot drat.
|
|
#
?
Mar 15, 2020 15:32
|
|
Have you seen my apex seals? I seem to have lost them.
|
 Some work on suspension geometry... pretty happy with it. Leaving out the graphs because they give away the game, but roll center moves almost exactly with the chassis under bump, doesn't migrate significantly under roll, camber curve is reasonable and motion ratio is pretty constant. Scrub is more than I'd like but that's unavoidable with Wide 5. Of course, in exchange for more scrub, I get hubs that are around 7 pounds and use big beefy bearings, and wheels that bolt to that which weigh around 7 pounds for a 15x7 and are about $200 apiece.
|
|
#
?
Mar 15, 2020 15:59
|
|
|
Uhhhhh, yeah those are trade offs that every single one of us would make. That's fantastic. Can I buy a spot on the shell once it's ready for a decal? I promise to keep it PG.
|
|
#
?
Mar 15, 2020 16:34
|
|
|
Yeah I mean I could do better by throwing a lot of money at things but enh. A while from there being bodywork but I'm perfectly happy to put hilarious things on in addition to the (undesigned, guess what's not my strength) livery.
|
|
#
?
Mar 15, 2020 16:49
|
|
|
 Just for grins... there's a certain amount of driver setback going on.
|
|
#
?
Mar 15, 2020 23:11
|
|
|
This is some extremely cool poo poo. Way more involved that I would mess with, but I’m not at all serious about racing. Still fun to watch (both the sport and the building...) Design and fab work is always interesting to me.
|
|
#
?
Mar 18, 2020 19:36
|
|
|
Darchangel posted:.... Way more involved that I would mess with... Funny, I'd say that about a lot of the restoration/etc work you did to the 86 and stuff. 
|
|
#
?
Mar 19, 2020 00:12
|
|
|
Oh, yeah. I mainly mean that my goals would be more street-oriented, so a whole swath of what you are doing isn’t just overkill, but counter to my use case, and I would put my efforts elsewhere. Fits your use just fine.  Design in service of the end goal and all.
|
|
#
?
Mar 26, 2020 20:50
|
|
|
What's that, a global pandemic isn't the best time to buy race car parts? Pfffft. Dry sump tank! 6" diameter, 30-some inches long, capacity about 3.5 gallons. Complete with heater. It's actually pretty rare to find tall/skinny ones like that, but they're the best at resisting uncovering the pickup from g-forces, and also packages more easily. Probably putting behind the "seatback" bulkhead on the "passenger" side, in front of the axle. Same seller has used quick change rear axle stuff. I need some stuff where it's not going to be common circle track options but dang if the price isn't right.
|
|
#
?
Mar 27, 2020 02:51
|
|
|
OK, I've been doing a lot of digging into shock stuff... you know, to educate myself on how to do things instead of paying an expert. As usual. Sigh. OK. So. 4 (or 7) post rigs are a bunch of hydraulic cylinders with servohydraulic valves, oriented vertically, where you place a car on top and wiggle stuff around (technical term). There's all sorts of things you can do, and booking time on them is fantastically expensive. Turns out there's a few hydraulic cylinders like that kicking around at work. But not 4. So thinking logically, I said, hey self, since you have this car designed and will have suspension designed shortly, what if you just made one corner? Make a full suspension for the left front corner, make a mockup to hold the pickup points in the right place, put linear bearings on that mockup and have it bolted to a structure to let it go up and down, and then put weights on the mockup side until the corner weight is right. There'd be a lot of stuff I could test for far far cheaper (you know, since my boss has no real problem with me using stuff on my own hours as long as I'm not interfering with anything) If you think about it, going over rough roads can be represented by random movement of the hydraulic platform and ride comfort (if this were a street car) could be looked at as a function of transmissibility... IE, you don't want vibrations going to the unsprung mass. Well, this isn't a street car but what if, say, you wanted to simulate going over kerbs or running on FIA kerbing and find something to make the car upset less and or tire have the least contact patch load variation? It just so happens that FIA kerbs (or any other style) aren't the hardest things in the world to simulate either. I mean that on top of a lot of other things to test that I won't bore you all with. So I think the first step is get a mule shock... I have a Penske triple that can go on the quarter car model, but I'm going to get a cheap(er) non-adjustable and a bunch of pistons to characterize things as a normal shock dyno first. I have first-guess shock curves figured out for the low speed side, but Penske piston designs can change how middle and high speed (shaft speed) portions of the graph react and I have guesses for what various options will do performance wise but guesses are worth dick and squat compared to data. So the idea is I get as many different piston designs together as I easily can, set them up on a different hydraulic thing acting as a shock dyno, come up with valve stack theories for every piston option, and then later run through those on the quarter car model. ... well, look, it's not like I'm going through the race car tire budget this year...
|
|
#
?
Mar 27, 2020 13:15
|
|
|
This is a reminder to myself to go through and post pictures, because I now have the non-adjustable Penske, and pile of various design shock pistons, and so on. But this week at work I'm going to finish welding up my chassis table so I can take it home this weekend. And then order a whole assload of steel tubing. Also designing my own tube notcher because some of the tube miter angles are on the crazy side for anything I can just buy.
|
|
#
?
Apr 28, 2020 23:34
|
|
|
mekilljoydammit posted:Also designing my own tube notcher because some of the tube miter angles are on the crazy side for anything I can just buy. Of course you are. 
|
|
#
?
Apr 29, 2020 19:49
|
|
|
Look, on the grand scheme of everything, it's not that bad. I just need a tool that doesn't exist at a price I'm willing to pay.
|
|
#
?
Apr 29, 2020 20:00
|
|
|
mekilljoydammit posted:Look, on the grand scheme of everything, it's not that bad. I just need a tool that doesn't exist at a price I'm willing to pay. I'd be stoked if you could provide more details about what you've got in mind and also where current options fall short. I've been looking into a tube notcher for thin-wall steel. I was particularly thinking about the Rogue VersaNotcher which looks great but is on the pricey side of notchers.
|
|
#
?
Apr 29, 2020 22:23
|
|
|
OK, so, when you do really shallow angles with hole saw based ones, the shallower the angle the deeper the hole saw needs to be. So there's guys that do custom deep holesaws where you cut off the back and weld a tube in. But it's still connected to a threaded that's like, 3/4"-10 or something and goes to a small diameter shaft. You pretty much need to for multiple sizes. I figure, screw it... designing for one dang size. So the holesaw gets welded to a, say, two foot long piece of tube the same diameter as the holesaw (1.5" in my case) with oilite bushings of the same diameter spaced maybe 6 inches apart or something. Weld a plate on the end with a hex drive for a drill, voila. Nice and rigid so it will drat well go straight. Only other real novel stuff I have in mind is technique... so you can set depth and rotate the tube axially cut several notches of different angles but having the same "crotch height" if you get what I mean. Like basically, for a given tube, you bandsaw cut it to the designed center-to-center length (ala the 3d sketch I have driving the tube setup in Solidworks) then fixture it such that everything fits after you're cutting the notches. It takes figuring out exactly what gets notched and everything else based on assembly order and stuff but, golly, that's easy to do now in digital.
|
|
#
?
Apr 29, 2020 23:00
|
|
|
I don't think I've seen a dry tank that tall even in buggies that do wheelies all day.
|
|
#
?
Apr 30, 2020 04:53
|
|
|
I feel like at a certain point it's easier to either: Back the hole saw out and hack the already-cut sections out of the way with an angle grinder, then go for it again Or Print a coping template on paper and wrap it around the tube, cut with angle grinder But I certainly am not in a position to throw any stones here when it comes to overcomplicating things.
|
|
#
?
Apr 30, 2020 06:46
|
|
|
Enh, the notcher sounds more complicated than it is. If I had the Bridgeport here instead of at my parent's I'd just use a rotary table and 1.5" roughing mill instead but I don't, so I'm going to get $50 or so of materials and make a notcher. I just hate having to cope with not-great tools, especially when I'm going to have to use it a lot.
|
|
#
?
Apr 30, 2020 13:25
|
|
|
wrong thread
Fermented Tinal fucked around with this message at 16:06 on Apr 30, 2020 |
|
#
?
Apr 30, 2020 15:45
|
|
|
Yikes it's been a busy couple of months - COVID has not meant any slowdown whatsoever at work so far. If anything the opposite, which is nice in a job security way but tiring. I swear I'll post pictures of cool stuff I got used at some point but not today. It's getting real when parts show up though. So first off, chassis table is almost done - just need to weld up a few legs and casters, then it'll be ready to truck home. Probably this week or next weekend. Goal is to get my first batch of steel and start cutting and welding tube - ideally on the 4th weekend which is a 4 day weekend for me. Second off... went and bought Matlab/Simulink/Simscape. And working through stuff but there's so many software packages that do what I want. Including multibody physics and stuff. See, I had the idea of doing a lap time simulation to iterate through changes a while back... turns out Simscape has most of the codebase needed to do it already written. And for that matter has tutorials on how to do exactly what I have in mind, albeit geared to Formula SAE folks. But hell, it's all adaptable. Today was trading emails with a guy at Calspan (https://www.calspan.com/services/transportation-testing-research-equipment/tire-performance-testing/motorsports-tire-testing/) regarding an RFQ. They may be willing to do things at a budget level I can swing; which would mean that my lap time sim would have actual real tire data for my actual real tires. Why no I haven't gone off the deep end, why do you ask?
|
|
#
?
Jun 22, 2020 23:12
|
|
|
Aaaaand I got an RFQ back. And it's actually pretty reasonable. I need to scrape together the money but holy poo poo I think that's happening.
|
|
#
?
Jun 25, 2020 13:24
|
|
|
 That would be amazing
|
|
#
?
Jun 26, 2020 22:45
|
|
|
I'm doin' this even if I have to sell stuff and postpone other stuff. It's too crazy and awesome not to. edit: Like, the tires for this class would be equivalent to 245/40/15 in normal sizing... but they're designed to work with class mandated 7" wide wheels. But because of that I have no confidence that any available tire data is anywhere close because they're so weird. So having test data? Hell yeah. They said there's a chance they could get two compounds in too.
|
|
#
?
Jun 26, 2020 23:23
|
|
|
I’m eager to see how the test data helps you out. Color me surprised at the relative affordability of making this happen.
|
|
#
?
Jun 27, 2020 03:08
|
|
|
That's really cool. Now I'm wondering if anyone has bothered to do this for other spec tires, say the options we have in club ford.
|
|
#
?
Jun 27, 2020 13:45
|
|
|
http://www.avonmotorsport.com/resource-centre/downloads has a lot.
|
|
#
?
Jun 27, 2020 14:47
|
|
|
OK, got some stuff done... Chassis table basic structure welded up and it's home now! Now just need to get steel.  Latest tube chassis version... and I think I'm good.  Been doing FEA on tube locations to try things... that's displacement graphed but stress is less pretty but more useful. I'm not going to go back and fix the geometry on the first try (in this thread a couple years back) but the very first version was sitting at 400 pounds. The previous peak was about 19,000 ft/lbs per degree of twist, and that version was 350 pounds. Now, tested the exact same way, it's sitting at 28,000 ft/lbs per degree of twist, and 250 pounds. Probably that will go up with paneling and stuff - I'll twist it on a fixture table when I'm done, for fun. But that's good enough.
|
|
#
?
Jul 18, 2020 14:22
|
|
|
Welp, chance ebay find means I ordered my flywheel today. It's not intended for a 13B. 
|
|
#
?
Jul 28, 2020 01:21
|
|
|
And in further escalation, think I got most of a quick change rear end on its way. Seller doesn't want to ship so trading favors with the guy who's buying my WRX
|
|
#
?
Jul 30, 2020 20:57
|
|
|
|
| # ? May 1, 2024 12:56 |
|
|
  So there we go. Right image is next to a stock diameter NA flywheel/clutch etc. Since I'm making a new bellhousing anyway, why base it off a big flywheel?
|
|
#
?
Aug 1, 2020 19:02
|
|