Does an LSD need better cooling?

[Zed Head]

Who will be doing the machining?  Have you estimated the total cost to completion?

Subaru still uses the R180 and there are a bunch of custom covers out there.  I'm not sure if the mounting stud dimensions are the same, can't remember.  I think that they are.

https://www.flatironstuning.com/rcm-r180-differential-cover

https://www.machinedspeed.com/product/custom-hi-capacity-r180-differential-cover

https://www.importimageracing.com/products/r180-rdf-asf-machine-r180-billet-rear-differential-cover-subaru-sti-2004-2021

[Zed Head]

2 hours ago, Jeff Berk said:

I was warned that they need to be tapered to allow them to release from the mold.

In sand-casting the sand is the mold.  It just brushes off.  The core, or form, is what might be molded, for high production parts, but they're not building a mold for $40.  Almost has to be 3D printed.

Interesting project, but getting all the way to a usable part will take many steps.  Good luck.

[Captain Obvious]

23 hours ago, Jeff Berk said:

 I have a spot on two of the fins where I can attach screws to mount an air deflector to direct air flow from below the differential across the back.

I was wondering what those bosses were for. I couldn't find any internal feature that would explain the necessity for those on the outside. I got it now.

So you were talking about tapering the fins so the part would release from the mold. In molding speak, they call that "draft", as in... "Are you sure you have enough draft on those surfaces so the part can be removed from the mold?"  I'm no mold expert, but it's my understanding that you only need draft if you are splitting the mold open into reusable pieces such as opening a die after die-casting. You aren't doing that with sand or investment casting. With sand and investment casting, you remove the mold in little pieces and don't need to salvage any of the original mold. And because of that destructive (to the mold) process, draft is not required. Now, it might help for other reasons like better feature definition or mold filling without air pockets or porosity, but I don't think you specifically need draft like you would with a die-casting.

As to the specific question... "How thin can I make the fins?" I would ask the foundry for guidance on that. And about the holes... I'm assuming you will be drilling all those in post-casting. You aren't going to to try to have any through holes cast right into the part, right?

And about accuracy of "as cast" surfaces, I'm assuming you'll be machining any critical dimensions into the finished part. If there's a sealing surface or bolt hole, etc, you'll be finish machining that and not counting of that coming out of the mold with enough accuracy. The thicker the material, the greater the shrinkage.

Lastly, have you talked to the foundry about the locations of gates and vents?

In my experience, things to consider... Locations for gates and vents for proper fill and prevention of trapped air bubbles. Uneven or too rapid cooling. Sink marks on thicker sections or where different cross section thickness intersect.  Porosity in the final part and how deep you need to machine precision surfaces. Warping or twisting of large parts.

Sorry for the novel. 

[Jeff Berk]

Thank you for all the input. My knowledge base for casting is at the low end as I’ve never had casting done for anything, but this is a chance for a new adventure of sorts.

I’ve modified the design based on comments, and I was planning  on printing a 3D test piece to send to the caster to review. I recall that he said to keep parts to ¼ inch or thicker and to have a draft as CO called it to make it easier to separate parts from the mold. I think that the concern might be that the separation of the 3D printed part from the sand to form the mold and not the separation of the cast part from the sand, but what do I know?

Besides external fins, I’ve also added fins to the interior to promote heat transfer between the fluid and the aluminum although aluminum to air is likely the more important thermal transfer concern. The interior fins should also add strength to the bulbous oil reservoir.

I was planning on doing all the drilling of the holes post casting.

With all the points CO brought up, I think this part is not going to happen for a few months while I research the process a bit more and communicate with the caster.

[Zed Head]

Looks like you added fins inside the cover.  Here's an interesting video from a guy who has a bunch more on the topic on his Youtube channel.  He's selling parts of course, but he does raise some interesting points.

 

[Zed Head]

How about remote cooling?  Like they use in laptops to cool the main processor.  

Borrowed this from a Youtube video.

[Jeff Berk]

I did look at the Banks video and borrowed some ideas from it like adding an under differential air deflector. His temperature studies also was the reason why I added a temperature sensor to the fill plug. I'm planning on using the original diff cover and observe the temperature during one or two Auto-X events. After that I need to change the oil and I was then going to try the new cover and see how it impacted the temperature measurements.

[Zed Head]

Have you thought about gluing cooling fins on to the cover you have?  Today's adhesives are very strong and can be conductive.  

https://www.3m.com/3M/en_US/p/d/b00006335/

[HS30-H]

Some quite hare-brained 'solutions' being offered here.

The original question was answered in period by the factory in their Race & Rally Preparation manuals. They recommended cooling of the oil by means of a pump and small heat exchanger, giving data for drilling the diff housing and rear cover for the insertion of tubes to collect the oil and then spray (cooled) oil on the CW&P:

 

 

A similar electric pump & cooler system was adopted as standard equipment on some of the S130-series variants and up into S31-series.

Rear covers with extra oil capacity and cooling fins are all well and good, but positive circulation, cooling and the directed spray of cooled oil are going to be more effective.

A lot depends on on how 'tight' your LSD setting is. Generally speaking, high breakaway settings generate more heat.

[Jeff Berk]

That adhesive is about $100. I actually have some conductive adhesive for attaching aluminum heat sinks to electronic components. 

I might give that a try and see if it helps during an Auto-X (and see if they stay attached).

[Zed Head]

I wonder how much heat is transmitted to the mustache bar?  A thought.  Also, what is air flow like around the diff?  Nissan's finned diffs put the fins on the bottom where the air flow is.  I wonder if anyone ducts air to the diff, like they do for brakes. 

It's an interesting topic to think about, especially with a low budget.  Have fun.

[Captain Obvious]

6 hours ago, Jeff Berk said:

 I think that the concern might be that the separation of the 3D printed part from the sand to form the mold and not the separation of the cast part from the sand

That's actually a very good thought. I've been unconcerned about getting the finished part out of the sand without draft, but getting the master part out of the sand before casting IS a concern. Any "vertical" surfaces where the master will have to slide against the sand should be avoided if possible. And the more sliding, the more chance you'll pull some sand off the wall and mess up the mold cavity. Like I said... I'm no casting expert! 

None of that would apply to an investment casting, but for sand made using a master packed in and then removed, it certainly could be.

So you need to look at the master and picture how it will be pulled out of the sand. Picture the parting line and plan the draft accordingly. The parting line is going to be right at the mounting surface, and your current design has several areas where you could probably benefit from some draft. For example, the fins around the bottom perimeter are tapered, but not in the direction that produces casting draft. Those fins are going to slide a couple inches scraping against the sand walls as the master is removed from the sand.

Any surface feature that is perpendicular to the parting line is suspect.