|
Big/little change... I have an idea, inspired by some modern TA iron, that may improve a lot of things chassis-wise. But fortunately it's all stuff that's not cut and welded yet anyway... and may also make some stuff easier to fab. Kid's in daycare so everyone in the house was sick for the last week but... getting back in gear.
|
#
?
Sep 30, 2021 15:33
|
|
|
|
| # ? May 1, 2024 23:17 |
|
|
 loving Mk15... but the latest changes do a lot for stiffness at a total penalty of like, 3 pounds. And no, it doesn't require any changes to the stuff I already have welded. Going to work on front suspension parts next because it's a fun change of pace. Oh, and plasma cut stuff for rear end will be done Real Soon Now.
|
|
#
?
Oct 7, 2021 00:08
|
|
|
mekilljoydammit posted:
What sort of steel are you using for this chassis? I'm curious about the type, OD and wall thickness for something like this.
|
|
#
?
Oct 7, 2021 18:12
|
|
|
So the safety structure is required to be minimum of 1.5x0.095" DOM mild steel. So it is. The rest is 1.5x0.065.
|
|
#
?
Oct 7, 2021 18:21
|
|
|
I started modelling front uprights and came to a realization that I'm a lot happier with... though it'll probably be a little more expensive it should be nigh bulletproof. Old design... red parts are aluminum. The hub contains the bearing and rotates on the spindle. This is fine and massively better than stock, but... I'm unsatisfied with the gussets - imagine there's a grease seal on the inside of the hub (forgot to put one into the assembly) and look how little gussets would add.  New design. Bearings are spread across wider area (more stiffness) and gussets have room to go out to the outer bearing too. And the brake rotor and hat can just come off. And it uses press-in studs instead of screwed in, and the bearings ride between steel and steel vs steel and aluminum, and... Anyway. A bit more costly to make probably but I think I like the tradeoffs.
|
|
#
?
Oct 9, 2021 16:58
|
|
|
You could even make your gussets with brake cooling in mind, or at least plan them so that your cooling ducts bolt conveniently to them instead of inconveniently.
|
|
#
?
Oct 9, 2021 17:02
|
|
|
kastein posted:You could even make your gussets with brake cooling in mind, or at least plan them so that your cooling ducts bolt conveniently to them instead of inconveniently. Sure can! Everything just packages better with the new design.
|
|
#
?
Oct 9, 2021 17:08
|
|
|
 Steering arm still needs to go on, but gusseting for the suspension and brake mounts is pretty satisfactory.
|
|
#
?
Oct 10, 2021 01:04
|
|
|
mekilljoydammit posted:So the safety structure is required to be minimum of 1.5x0.095" DOM mild steel. So it is. The rest is 1.5x0.065. Thanks! Is there a reason cromoly, like 4130, isn't used? Also, what filler are you using, something like ER70S-2?
|
|
#
?
Oct 12, 2021 19:08
|
|
|
casque posted:Thanks! Is there a reason cromoly, like 4130, isn't used? Basically I'm stiffness limited not strength limited - 4130 normalized would get a touch more ultimate tensile strength (not a limiting factor) no extra stiffness, potential pitfalls with welding, and twice the cost. Edit: that's true for the chassis though - suspension uprights will be 4130 and then heat treated. And yeah, 70S2.
|
|
#
?
Oct 12, 2021 20:08
|
|
|
Isn't it way lighter per foot used vs regular steel? Local guy that runs a drag shop keeps trying to talk me into doing chromoly cages and I keep shooting him down for lack of experience and heat treat capabilities.
|
|
#
?
Oct 12, 2021 23:40
|
|
|
honda whisperer posted:Isn't it way lighter per foot used vs regular steel? Local guy that runs a drag shop keeps trying to talk me into doing chromoly cages and I keep shooting him down for lack of experience and heat treat capabilities. Only if you go thinner wall; density is for all practical purposes identical. Some drag racing rule books (at one time? IDK) allowed thinner wall cages in 4130 vs mild steel so that might be where they're coming from.
|
|
#
?
Oct 13, 2021 01:07
|
|
|
mekilljoydammit posted:Only if you go thinner wall; density is for all practical purposes identical. Some drag racing rule books (at one time? IDK) allowed thinner wall cages in 4130 vs mild steel so that might be where they're coming from. Yep that's it. Digging through the NHRA rulebook it's almost 1/2 the wall thickness. NASA is 1:1. I assume NASA is cribbing the SCCA notes. I've only done one NHRA roll bar and it was years ago so very rusty on their rules.
|
|
#
?
Oct 13, 2021 01:30
|
|
|
mekilljoydammit posted:Basically I'm stiffness limited not strength limited - 4130 normalized would get a touch more ultimate tensile strength (not a limiting factor) no extra stiffness, potential pitfalls with welding, and twice the cost. Thanks, this is a really cool project you've got going. What are the pitfalls of welding 4130?
|
|
#
?
Oct 13, 2021 04:05
|
|
|
casque posted:Thanks, this is a really cool project you've got going. What are the pitfalls of welding 4130? So in the grand scheme of heat treatable steels it's not that bad... but the conditions you'll get it in are annealed or normalized. Annealed is not stronger than mild steel, normalized is. Now any sort of heat treating of this stuff is done by heating it up to 1600F or so and then letting it cool at various rates... too fast and it's hard and brittle and may crack, too slow and it's annealed. In the old days they used oxy/acetylene for welding this stuff which had a wider heat affected zone and meant that more stuff was hotter so it all cooled a bit slower; TIG is much more concentrated heat input so the part that got really hot cools faster, and post-weld-flow of argon is... well it's a cool gas blowing on stuff. The accepted way to re-normalize everything is use heat indicating crayon, heat it up to 1600F and let it cool in still air. So it's not rocket surgery, but there's extra stuff you have to be careful of with how you put heat in and what you do after welding... as opposed to mild steel where you don't have to do squat. For the suspension uprights I'm going to be heat treating, I'm actually looking into a used lab-style hotplate to just leave everything on while finish welding to preheat and control the cooling rate.
|
|
#
?
Oct 13, 2021 12:45
|
|
|
When you say DOM mild steel, do you mean something like 1018 for grade? I'm getting ready for a welding project of my own and chasing grade and manufacturing process with suppliers has been a bit confusing for me. I'm assuming the cages can be thinner with 4130 because with higher UTS at similar stiffness you get more energy absorbed before the material fails? I've not done a tone of preheat welding (or much welding at all tbh) but I've had good luck with annealing hardened steel parts with a normal hot plate, cast iron skillet filled with sand and a thermocouple in it to monitor temp. It's not closed loop but pretty easy setup for keeping parts at elevated temp for a long time at a low cost. You might have mentioned this but youre gonna tig the frame? I was struggling with not burning through tubes unless fitup was near perfect so I didn't have to fill any gaps, but again I'm a pretty amateur welder. E: are you using SW FEA tools for checking stiffness? Do you have targets for flex at load or just doing relative comparisons between designs to decide on bracing strategies? Still enjoying following along and enjoying watching the progress.
|
|
#
?
Oct 13, 2021 19:43
|
|
|
They make A36 DOM, right?
|
|
#
?
Oct 13, 2021 20:18
|
|
|
Yeah I'm struggling to separate out manufacture specs from grade of steel from shorthand names. I'm planning on using square and rectangular tubing instead of round which adds another layer. I think I want a513 mechanical tubing, but that seems to cover multiple steel grades and when it's a513 round its considered DOM but manufacturing method isn't listed for square or rectangular tubes so I'm confused.
|
|
#
?
Oct 13, 2021 20:52
|
|
|
meowmeowmeowmeow posted:When you say DOM mild steel, do you mean something like 1018 for grade? I'm getting ready for a welding project of my own and chasing grade and manufacturing process with suppliers has been a bit confusing for me. The stuff I'm getting is A513 Type 5 DOM, 1020 in the size I'm using. I think the rulesmakers are making that assumption about energy absorbtion or whatever, but looking at the way rulesmakers have gone in eliminating the thinner wall allowance for chromoly, maybe it's proven to not be the case. It may be a bit blase of me but I'm a bit at... the rules specify stuff for cage, I build it to that and move on with my life. A lot of these cars have gone upside-down or whatnot over the years, the specified cage construction seems relatively safe... voila. That's actually a really good idea with using a skillet full of sand... there's some cheap closed loop controllers I could just have read off the thermocouple too. I think I'm going to try that. I am TIGing the frame; and really I'm just being really really anal about fitup. Makes everything take longer but with some of the measuring mitering techniques I showed I've gotten it to where usually tubes are a bit of "need a push to get to the desired position, joint with no obvious gaps". But there's a number of tubes that were intended to be long tubes that since became used for shorter tubes, heh heh. FEA... well. Basically I'm not sure how much I trust Solidworks for absolute numbers - and anyway, there's no paneling or a bunch of other stuff. So yeah exactly, I'm basically doing A-B comparisons, on the order of "ooh I had an idea, let's see if it sucks"
|
|
#
?
Oct 14, 2021 00:13
|
|
|
Thanks for the info, I'll probably also go a513 and just see what it looks like. Sucks that our local steel yard here got purchased by standard steel a year or two ago and their online catalog disappeared so you either have to call or fill out a quote request with exactly what you want so its not great for browsing to help decide what you want haha. And yeah I'd take the same approach with frame and cage design, unless you have the tools to prove out something is better you might as well go with the spec or standard for the class. I've done the closed loop controller thing for some stuff and yeah that works, I can't remember who made it but I have a controller at work that plugs into the wall and has a thermocouple port and a 3-prong port, dial the temp and it PIDs the output to hold it there. Super convenient and maybe cheap? I've also run an old oven via PID but that was with an omega panel mount PID controller running a relay that switched the hot line, worked good enough for some carbon fiber stuff without building a full-on autoclave/oven jobber. Half the reason I'm doing square and rectangular tubing is to simplify the cutting, I've got a cold saw I can use that'll do brilliant cuts as long as it doesnt need to match a round tube. If I was gonna do round tube I was thinking of doing something kinda like what you are with the collars but using a toolroom lathe to spin the hole saw and attaching the tube to the toolpost but I realized I wasnt getting much from round tube as everything it was mounting to was flat.
|
|
#
?
Oct 14, 2021 01:48
|
|
|
If I was doing this thing as anything other than a labor of love I'd either do it in square tube, or figure out a plasma/laser CNC tube notcher. I'm actually going to bang out a copy of a proven Trans Am design (solidworks weldments make that relatively fast) in order to compare it to my stuff just as a sanity check. Problem with spinning the holesaw in a lathe would be that you wouldn't be able to cut very deep - on some of my shallow angle miters I have to pause halfway through the cut and cut some of the tube out because the holesaw is bottoming out, gripping it in a lathe chuck would sacrifice some depth. On the other hand it would be marvelously stiff...
|
|
#
?
Oct 14, 2021 01:54
|
|
|
Why do you see round tube as the easier option? I guess with compound miter joints its harder to do the cuts than round tube with your system, and tbh that's a concern I have. I have a couple of compound cuts to make and tbh I'm not sure how I'm gonna do them, I don't think my cold saw has a compound miter adjustment. I guess if thats correct, what's the perceived benefit of square tube? Most automotive stuff I see is round tube so I figured it was standard. SW weldment tool is so close to ideal, I just wish you could choose to align profiles at both ends of a line segment instead of one. On my design I have one plane of the structure that needs to fit inside of something so I'm using the sketch to define the outside of the profile and the other plane has something that needs to go into it, so the sketch marks the inside. Theres no clean way to get the connection pieces to fit correctly that I've found, I've had to draw up a separate sketch that goes from the two tube sets centerlines to get it to all line up. I was gonna turn a solid extender shaft that would hold the holesaw out from the jaws to clear. I had an annular insert cutter that was like .002 undersized on one of my tube ODs and would make snap fit joints, it was magical. Are you shopping out the milled and turned parts in the suspension or doing yourself?
|
|
#
?
Oct 14, 2021 03:26
|
|
|
Round tube is 100% not the easier option in any way; I could be mitering basically everything with a horizontal bandsaw if I were building out of square tube. Formula car chassis and stuff tend to be round tube, but bigger stuff like Trans Am cars or other GT stuff tends to be square tube. I'm planning on DIYing the machining - I have a mill and lathe. If I ruin too many parts I'll reconsider though.
|
|
#
?
Oct 14, 2021 12:35
|
|
|
Getting back in the swing of things... new barn doors on barn are as-done-as-they're-going-to-be this season so can get back to the chassis. And got trained on the Zeiss blue light scanner at work.  So this is a scan of an EMP Stewart Pro Series water pump cartridge - been intending to do this for purpose of making a new housing for rotary use. Can't see it on the crummy phone-photo-of-work-laptop-screen but the scan resolution is high enough detail to see machining marks. Tomorrow I bring in a wheel and one of those PFC calipers so I can do proper, accurate mockup. Over the weekend I'm bead blasting a couple front covers and scanning those next week... the QC department at work's factory (in Mexico) is eventually going to get the scanner now that engineering isn't using it as much, so want to get a bunch of important-to-my-purposes stuff done ASAP.
|
|
#
?
Nov 19, 2021 00:53
|
|
|
That is so cool. How many viable reasons for keeping this device in engineering have you managed to invent so far?
|
|
#
?
Nov 19, 2021 01:08
|
|
|
Seconding, that is insanely cool. The fact that we can scan things in such minute detail and hell, make three-dimensional prints of them ... living in the future is pretty cool.
|
|
#
?
Nov 19, 2021 01:49
|
|
|
taqueso posted:That is so cool. How many viable reasons for keeping this device in engineering have you managed to invent so far? Well, they don't have anyone qualified to use it in the factory so I've got a bit but... I'd much rather just bang out a bunch of stuff on Thanksgiving week when nobody much is around anyway than risk it going away. It might not simply because we kinda don't trust the QC reports the factory gives us with calipers so this would just be more data we don't trust... but that's not my call.
|
|
#
?
Nov 19, 2021 01:58
|
|
|
mekilljoydammit posted:Well, they don't have anyone qualified to use it in the factory so I've got a bit but... I'd much rather just bang out a bunch of stuff on Thanksgiving week when nobody much is around anyway than risk it going away. Ah Thanksgiving week in Wisconsin when everyone is out hunting, good time to get stuff done without being bothered.
|
|
#
?
Nov 19, 2021 14:56
|
|
|
Latest on the turntable, a PFC ZR34! Great calipers honestly; compatible with the Wilwood superlite, but monobloc Look at that sucker; OK it's not new but still good.  Screen picture of what the scanner is doing; I like how it looks almost arty  And after a few scans, that's enough data to reverse engineer. Or at least it has all the critical points where I can plunk it into CAD to figure out how to fit it into the wheels.
|
|
#
?
Nov 20, 2021 05:44
|
|
|
  And now a 12A front cover; I intend to reverse engineer both this and a 13B-REW one and make a mashup where the REW timing setup replaces the 12A distributor hole, but with the 12A motor mounts and stuff.
|
|
#
?
Nov 23, 2021 01:23
|
|
|
This is super cool. Prints of the scans would make really cool garage art too!
|
|
#
?
Nov 23, 2021 01:37
|
|
|
 And there's the REW front cover. Handling the reverse engineering into CAD part will be a pain unless I figure out a better way, but now I have dimensionally accurate data I can pick at to get there.
|
|
#
?
Nov 24, 2021 01:30
|
|
|
Yeah, rebuilding a casting in CAD is always a hoot
|
|
#
?
Nov 24, 2021 01:49
|
|
|
It helps that I'm a bit... like, there's design points I need to maintain (mount points, bolt holes, timing mount points, clearances etc) but if I don't copy the surfaces exactly, I am not obligated to care.
|
|
#
?
Nov 24, 2021 02:18
|
|
|
Yeah, that should make things a lot easier at least. Just gotta make sure you maintain wall thickness, draft, oil drainage, and clearance to any stuff the part covers I guess. How do you account for shrinkage? Is it mostly isometric or will parts shrink more in certain directions?
|
|
#
?
Nov 24, 2021 04:04
|
|
|
So far from testing it seems isometric-enough and about 1.5%; I do not guarantee that that's correct but it seems to work so far. The great thing about having real parts, especially when they're scanned like that, is that I can crib solutions to things... like these were parts that were clearly mass produced, from some of the geometry quirks I can figure out how and things like drafts. Or just try stuff, make junk parts, try to figure out why, and remelt the aluminum.
|
|
#
?
Nov 24, 2021 04:19
|
|
|
Using the NADCA design guide charts for draft and fillet size can make things a bit easier too, even though die casting requirements aren't always the same as sand casting. You can probably just add like 1.5-2 deg of draft and make the fillet radius the same as the wall thickness and be done with it. This is also super rad and I wish I had access to nice scanning equipment.
|
|
#
?
Nov 24, 2021 16:48
|
|
|
  Wheel scanned... very rough but have complete coverage of what I need to reverse engineer, so good enough.
|
|
#
?
Nov 25, 2021 00:56
|
|
|
Very cool.
|
|
#
?
Nov 29, 2021 14:08
|
|
|
|
| # ? May 1, 2024 23:17 |
|
|
Sitting and chugging away at getting bodywork modeled - so I want to be able to do this again for more cars so I'm trying to get a whole toolchain of making a bodywork plug from a CAD model. Which is not the simplest thing, go figure. RL has had a lot of stuff, let's say; spring may be a bit less stressy. In the mean time also working on an idea for the larger class... see the tube chassis class is having participation issues... part of it is people parking cars and stuff, but I want to help ensure long term survival. One of the ideas I had is putting together a build manual of ... basically like the classic Locost books, with plans and everything, except more focused. "Here's plans on how to build an SCCA GT3 chassis roller, starting points for suspension including both geometry and shock/springs, find someone who makes bodywork, plunk your driveline in and you now have a starting point car that's upgradeable." So contrary to the complexity of my passion project RX-7, it's all square tube and easy tube miters... I'll still be doing FEA to do a good job on stiffness but the emphasis is "no harder than a Locost" for the chassis structure.
|
|
#
?
Dec 10, 2021 14:45
|
|