1/16 Scale FrankenDiff

[ddmckee54]

I'm going to be doing several Bruder conversions in the near future, since I'm retiring next week. I don't have the skill to scratch build, and I like the idea of using the Bruder stuff as a starting point. Bruder at 1/16 scale really doesn't look right when you use 1/14 scale running gear, plus the cost of the running gear is more than I'm willing to spend.

Soooo... my wandering mind latched onto the idea of Frankensteining a reasonable looking truck rear end out of other available low-cost bits. I probably got the idea from "mick thebass" on YouTube when he modified a couple of low-buck 1/12 scale WPL diffs and used them on his articulated Bruder dumper. He did some really neat stuff if you haven't checked him out. He hasn't updated his channel for a couple of years so I'm afraid he might not be with us anymore - which is too bad. The guy was a master at stuffing 10 pounds into a 5 pound sack.

I thought about using the Kong 1/12 scale diff, even got one to see if it'd work. It's a GREAT looking axle, smooth as silk, about the right size, and the price isn't too bad. But as far as I know they only make the rear axle version, they don't make a center axle version. I don't know about you guys, but the thought of building a truck that's SUPPOSED to have double drivers, and then having one axle sitting there doing nothing doesn't appeal to me.

So, back to the Frankendiff. I'm starting with your basic plastic diff for a WPL 1/12 scale car, it's guts are about the same size as the Kong axle. The actual axles will need to be shortened, but that'll be fairly simple machining to cut them to length, turn down the ends, thread them M4, and drill new holes for the drive pins. I'm just going leave it an open diff for now, no lockers.

Nope, the complicated part is the 3D modelling of the front and rear halves of the differential housing so that the WPL guts will fit into them. (While simultaneously learning Alibre, I use-ta-use Designspark3D.) The pinion housing is the most complicated part and I re-did that last night - I was NOT happy with the first attempt. I think I should be able to get the rear cover done tonight, it's pretty simple.

Then I'll 3D print a prototype and see what parts need to be fixed. I'll take some pictures when I've got some physical parts.

Eventually I'd like to cast the parts in ZA-12, both for durability and the extra weight. (But plastic will do for now.)

[Zabco]

Welcome to the good life. I was able to take early retirement 8 years ago and I have enjoyed every minute of it. What got me into this hobby was the fear of 'what am I going to do with all this time' after averaging 50+ hours a week for the last 30 years. Has not been a problem. Get to do everything I like to do, research, design, build, play with electronics, play with machining, experiment with new methods, etc, etc, etc. Most weeks, come Friday, I wonder where the time went.

See you apparently decided to go with Alibre as your 3d cad program. It's been a good program for me, hope you enjoy it.

[ddmckee54]

Zabco:

I went with Alibre for a couple of reasons, you can get a perpetual license, and you can get an off-line version of that license - one that doesn't need to phone-home every 30 days. I refuse to rent a cad package, and I also refuse to let somebody else hold my files hostage. I pulled the trigger on the Alibre purchase this morning, only had 4 days left on the trial. I've REALLY got to learn Best Practices, and actually use them, instead of just poking and prodding at it like I have been.

I printed out a prototype of the front half of the diff to see how things fit. Let's just say that my measurements were apparently less than accurate. Or at least my 3D model is less than accurate. I guess I got some 'slpainin to do to it. Over the weekend I hope to get to the point where the WPL guts will fit between the front and rear halves of the diff.

I left side to side clearance for fit of the pinion to ring gear, as well as front to back clearance for the pinion to ring gear. I figured I'd print a prototype and let the diff find its' happy spot, then measure how much I needed to adjust the model.

I've definitely got a problem with the diff carrier bearing fit - they don't. Looks like I measured the bearing ID and OD, then drew the 3D model using the ID dimension instead of the OD.

Then I've got this STUFF in the middle of the axle itself, that I can't for the life of me figure out WHY it's there. On both axles there a hole with a very short pin inserted in it, imagine a 6mm long pin through a 5mm axle and you've got the picture. I'd include a picture, but that axle is at home and I'm NOT. There's a plastic collar that slips over the pin and retains it, but the pin also limits how close the outboard edge of the collar can get to the diff carrier. Just outboard of that junk is a plain bushing, yup - a bushing. There are bulkheads on both sides of the bushing fixing it in place. The inboard bulkhead also does a very rough job of limiting the side to side position of the diff assembly by limiting how far outboard that plastic collar can move.

I can't think of ANY good reasons why the crap is there. The best explanation I can think of is that WPL uses this assembly in other differentials and it's needed in that other application. We all know that WPL uses bits and pieces from this model and that, slaps them together, and calls them a new model. In the next round of design I'm going to chuck all that crap in the middle of the axle into the trash-can and see if anything breaks. I've still got bearings supporting the diff carrier and bearings on the outboard ends of the axles. That's all that most full-size cars and light trucks have. If it's good enough for them it's good enough for me.

It's supposed to be rainy Saturday night and Sunday, we'll see what I can get done on the Frankendiff.

[Zabco]

Same reasons I went with Alibre also.

Yep, bearing support for axle shafts in middle and end should be only thing you need. That's all the Tamiya rear ends have and they seem to hold up just fine to all sorts of abuse.

[ddmckee54]

I got some stuff printed out over the weekend. It's been a while since I created a new album, I THINK I've got it set for sharing, let me know if I remembered all the hoops I need to jump through so that anybody else can see the photo.
https://photos.google.com/share/AF1Q...l0RnQ5b3hWdnVR

What you are hopefully seeing is the front and rear halves of the diff housing screwed together, along with the pinion housing screwed on the front. I've lost track of the number of times I've printed the diff halves, definitely more than I can count on one hand, even if I was the 6-fingered man. I was having bed adhesion problems at first, it's been a while since I printed anything. But I DID dust the bed off first, that just wasn't enough. That was Round 1.

Then it JUST wasn't measuring up, too small. That was Round 2. Yeah, I forgot to scale it up in the slicer - it's been a while since I printed anything. I scaled it to 103% in the slicer and printed Round 3. Hoo-Boy-Howdie Round 3 measured at 120.18mm, close enough to 120mm for me. Do the guts fit....NOPE. How come, the ring gear has a 25mm OD and it's in a 26mm hole, it should fit. IF the ring gear was centered in the opening it would fit, but when the ring gear is offset from center - not so much.

Re-design the diff carrier space in the front and rear diff halves and re-print - Round 4. Things were looking up after Round 4, outside dimensions checked out OK, the donor diff guts and bearings appeared to fit in their respective pockets. Will the front and rear halves fit together? NO, the diff halves do not fit together!!!!!

I started measuring the bearing pockets, since the bearings are the only things that could be keeping the case halves from mating together. The bearing measurements and the model measurements match. I don't like either of the explanations that I was able to come up with; 1) either the slicer is doing something funky when it's scaling the part, or 2) the part isn't shrinking linearly.

The only thing I could think of was to start tweaking bearing pocket dimensions and see if that worked:
Round 5 - too small.
Round 6 - better but still too small.
Round 7 - WHOOOPS - over-compensated.
Round 8 - Sorta worked. The diff halves do clamp together, but the axles not longer turn as easily as they used to turn. I wondered if I was clamping the bearings too tightly? I backed off the screws and sure enough things started turning easier. It's hard to believe that a pair of M2 sheet metal screws and a 2 piece plastic clamp could apply enough pressure to load the bearing shells like that. I'll tweak the pocket dimensions one more time and try Round 9 tonight.

Don

[Zabco]

When you create the .stl file from your 3d model in alibre a stl configurations box should pop up. Make sure and use a value no greater than 1 degree for the faceting parameters setting. stls do not contain true circles, only straight lines. A value of 1 degree would create a circle comprised of 360 straight lines. If your faceting was set to say 6 degrees then the circle would only consist of 60 lines. The lines would be longer but more importantly the center distance between two opposing lines (the diameter of your circle) would be less than the desired diameter. This would compress your bearing when you tightened up the two axle halves.

[Blender]

Awesome you took the CAD plunge. Alibre was at the top of my list too.


Id wager your printer is culprit. Some are better than others. Should be able to tune scaling in each direction. I've never totally solved it. Bearing pockets need a pretty tight fit. The slight compression of each layer squeezes it a bit wider.

3D printing is certainly a bit more involved than push button, receive part.

Is it possible to assemble the axle halves, and thru drill/ream the bearing pockets? Or are the end opening too small?

[frizzen]

With your growing lack of tolerance for fighting the printed part tolerances: What if you rough assembled the axle guts, moderately warmed them with a heatgun, and tried to close the axle case halves around them? Let the parts soften and clearance what they need, then you can evaluate where it made marks.

Which style WPL axles actually have a Differential instead of Spool?
I had wondered how much the extra bearings in the wpl axles was there to stiffen up the cases instead of support the axles?

[ddmckee54]

Nice discussion we've got going on here, deeply appreciated.

Blender:
I use Simplify3D for my slicer and I can scale all the axis independently if I want. I've tweaked the basic setup of the printer, it's a Wanhao D6 clone by Monoprice, so that I get accurate external dimensions. The diff halves should measure 120mm across, the circular part of the housing should have a 37mm OD and it should be 7mm thick, they measure 120.18mm, 36.98mm, and 7.21mm respectively. The first 2 are close enough for me. I'd like to get the Z height a little closer, but since my version of S3D does not allow for variable layer thickness, I've got to accept that the thickness could be off by a layer thickness, 0.2mm is what I'm currently printing at. External is fine, it's the internal pockets that are the problem. Drilling/reaming the 9mm outer bearing race pockets wouldn't be a problem. The only way I could see to get to the inner 12mm pockets would be some sort of line boring set-up, and I've only got the 6mm hole for the axle to work through.

Zabco:
I've been using the default set-up for the STL's, which I think is 4 degrees. Since I've been poking and prodding at all 4 of the parts, and have to reprint them all, I'll look at making a higher resolution STL.

Frizzen:
This is a replacement WPL diff for the "1:16" scale trucks.
https://www.ebay.com/itm/22618175378...Bk9SR-DG5t3EZA

I don't remember the exact part number, but this is REALLY similar to what I am using for a donor axle.
https://www.ebay.com/itm/20485224934...3ABFBMkKPg3cRk

[ddmckee54]

Two steps forward, one step sideways, and one step back.

I measured all of the internal pockets on the diff halves, they were ALL consistently about 97.4% of the model size. That's the size I would expect them to be when you allow for shrinkage. I'm beginning to wonder if my version of S3D scales internal dimensions? I really need to try scaling the slicer to about 150% and see what the pockets look like then as an experiment - without printing anything.

Anywho... I scaled the critical pocket dimensions to 103% in the model, the same as my scaling factor in S3D and printed the diff halves again - things are REAL close. The inner bearing pockets are a little bit too tight, and the outer bearing pockets are a little too loose. When I tighten all the screws holding the halves together the outer bearings will slip out of the pocket, and the inner bearings will not turn freely - but I think that's MY fault.

The outer bearing pocket dimension when scaled was 4.84mm and change, I just rounded it up to 4.85mm. The inner bearing pocket dimension when scaled was something like 6.12mm and change, I just rounded it down to 6.1mm. I'm going to print this one more time, but this time I'm going to round the pocket dimensions to the 3rd decimal place and see how that works.

Don