My EFI Conversion

[the_engineers]

In hindsight, the problems with the first pour were as follows:

1. The floor was too thick (1/2"), consuming too much metal.
2. The throttle body mounting flange was too thick (1 1/2") consuming too much metal.
3. The long, thin expanses of the side walls were too weak to survive the ramming process and deformed, appearing as warpage.

So, taking these lessons to heart, the 2nd mold was designed and constructed.

Additional webbing was added, as well as a "bridge" to tie the two side walls together for better strength in the casting process.



The floor of the mold was created with 1/4" material and tripled-up at the mating surfaces



Internally, webbing was added for additional strength.  These webs will remain in place until they are proven (or disproven) through future testing.  More than likely, they will be cut out later.





I jumped the gun a little and cut this hole to conserve aluminum.  Unfortunately, it's too high.



However, it's easily fixed...



Due to weather (and the spacial constraints of the foundry), the pour has been pushed back to Friday.  The extra time did allow me to come up with 25 lbs of automotive grade aluminum (2002 Hyundai Sonata wheels), though.

Also, I feel remiss in not providing this sooner, but the foundry doing my plenum is Working Metal in Urbana, Illinois.  The owner/proprietor is Dean Rose and his website is http://www.workingmetal.com.  Dean is a metal magician and keeps himself occupied with special requests and projects from the local community, the University of Illinois and Carle Foundation Hospital to name a few.

If you have anything you'd like to see in aluminum, feel free to contact him directly or work through me if you need a form built.

[Grundy]

That looks very cool.

[ksierens]

Wow, you are doing a great job.  I have worked on patterns in the past, and I remember using some type of max to smooth out the surfaces also.

[PlumCrazyChris]

Have you considered submerging the form to determine the exact volume of your pour?   Sounds like you have it nailed now, but it would be a good check.

Finding the right vessel to do it in (straight sides, etc.) may be a problem, but it would be doable.  I'd add enought water to cover the form, stick my hand in to my  wrist and then mark the water level.   Then push the form under and mark again.   Measure the difference and calculate the volume.   

Then you'd use a conversion factor to come up with the cold wt. of Al you need, which you probably know already.   

Just wondering, is there much of a volume difference between cold and molten aluminum?

Then maybe add a couple of percent to ensure a good fill.... 

I love this process, great work, I'm really looking forward to seeing your finished product.   

Great thread, it should be archived.

Chris

[the_engineers]

Have you considered submerging the form to determine the exact volume of your pour?   Sounds like you have it nailed now, but it would be a good check.

We actually talked about that.  Right now, I'm relying on AutoCad's Solid Object property calculator, which puts the volume at 137 in³.  The weight of Aluminum is .098 lbs/in³, so I should be looking at 13.43 lbs of plenum.  I may do a little to reduce this such as reducing the thickness of the bridge.

Aluminum does expand, although I'm relying on his expertise for that.

Looking forward to Friday 

[BruteForce]

The aluminum will shrink a bit, so you'll need to make the form oversize. You can calculate it by taking the coefficient of thermal expansion for aluminum, and multiplying it by the difference between the melting point and room temperature.

(660 - 20) * .000023 = .015

So, it'll shrink 1.5%.

[miketyler]

The aluminum will shrink a bit, so you'll need to make the form oversize. You can calculate it by taking the coefficient of thermal expansion for aluminum, and multiplying it by the difference between the melting point and room temperature.

(660 - 20) * .000023 = .015

So, it'll shrink 1.5%.

I just think its cool that you knew that. 

[the_engineers]

The aluminum will shrink a bit, so you'll need to make the form oversize. You can calculate it by taking the coefficient of thermal expansion for aluminum, and multiplying it by the difference between the melting point and room temperature.

(660 - 20) * .000023 = .015

So, it'll shrink 1.5%.

So can I minimize the shrinkage by lengthening the cooling time for a longer period?

[BruteForce]

So can I minimize the shrinkage by lengthening the cooling time for a longer period?

No, the time period has nothing to do with it. Just the temperature difference and the coefficient of thermal expansion. (I looked these numbers up on Google, no, I don't have them in my head  )

You can verify the shrinkage by checking your first casting dimensions against your measurements.

[the_engineers]

Just got done with the 2nd attempt.  Learned a lot, however it looks like a 3rd try will be required.  The fragile nature of the foam is both a blessing and curse with this casting process.

I got a chance to actually get my hands dirty this time!

The first few pictures are of ramming the form.  Basically, this involves burying the form in well-packed casting sand.







This will be the feeder to the rest of the casting.  The inlet vent will feed directly to this piece.



Once the bottom side is well packed, a screed is run across to level the sand.  A bottom board is then placed on the form and the assembly is flipped as a unit.





The top is then filled in and packed the same as the bottom.



Vents are added to let the aluminum in and (hopefully) out.





While the form was being rammed, Dean was adding aluminum as additional pieces would fit.



Finally, the form and the aluminum were both ready.  The real mystery in this process is the pour itself.  You spend the time hoping that all is well.



Ten minutes after the pour, it's ready to see.  The sand is bonded with motor oil, hence the massive amounts of smoke.







The minor repair that I did to the throttle body mounting flange held up well.


We're getting closer.  We're still short on material, but very minimally now.  The unsupported areas still ended up wider than the bridged section.  The next version will have more bridging to help minimize this issue.





I'm also going to do some reading about investment casting.  Perhaps the ramming process is too hard on a form with this level of shape complexity.

[the_engineers]

Knowing that I was continuing to come up short on material, I started really analyzing what was wrong with the forms I had built to date and how I could strengthen the top rail without adding material to the overall casting.

As such, the third (and hopefully final) design incorporates a rail design similar to an oil pan rail...thicker and broader than in previous designs.  Hopefully, this 1/2" thick x 3/4" wide section will prevent warpage during the ramming process.

To offset the extra material consumed in the rail, the stantions and supports throughout the rest of the casting have been removed.  All of my eggs are in this basket.

Also, after being pretty unhappy about how the lid was turning out in wood, I went ahead and made it in foam as well.  We've seen remarkable clarity and detail with the foam, so I think some of my early misgivings were just that.

Here is the latest version...








Again, I'm running the black cast Mopar Performance valve covers, so I've got a general theme I'm shooting for.

Cross your fingers and wish me luck.  We pour tomorrow at 4pm CST and I'll have pics to follow!

[hemiken]

It will be awesome, if i could make a suggestion..... Maybe the volume of sand can be controlled, by using a thin piece of ply on each external side of the plenum box as a support for the weight of the sand, just to control warpage when ramming, you could use it as a false wall of sorts.  And i would still use a layer of sand in between the foam and ply support wall to get the desired sand casting look on the plenum box rather than a super smooth finish that wood would give you.

[the_engineers]

I like that idea a lot.  It'll get painted, so I'm not as worried about the appearance.  Worst case, I glue 120 grit sandpaper to the plywood first...

Need to think about this some more.  Thanks Hemiken!

[hemiken]

  No worries.

Post up what you decide to do and plenty of pictures........

[r15303]