Sump! (or oil pan if you prefer)

Poke while the pattern is still in but stop short of the pattern. Be generous about it. If you touch the pattern youll get unwanted wires, and if you do it with no pattern youll knock chunks of sand out.

EDIT: I didnt see Kelly's reply above till I'd already posted. I wouldnt have just blatantly contradicted him. Not kissin ass, just sayin. My own experience does support my comment though.

 

Hokay, so having been doing a bit more inspection, I have found that:

1) My horizontal surfaces are somewhere between 1.5 and 2mm thicker than designed. No wonder it ate all my metal! I think the dimples between the fins are shrink defects due to this being so much thicker than planned.
2) There's a really nasty shrink defect near one of the bosses in the corners that are thick part, in an island of thinner bits. If there isn't an actual hole, it must be only a foil-thick bit of metal separating inside from outside.
3) I have clearly thought very poorly about section thickness and order of freezing, with my sprue and riser really providing very little help.

So while I think this casting may be usable with welding, I think I'm going to do the following and have another go:

- Sand the underneath of the cope pattern to lower it by ~1.5mm
- Redo my feeds as follows:


I'll pour into a 22mm hole that feeds a 16mm thick slab that in turn feeds the 10mm boss that had the worst of the shrink defects in it (plus sneaking it onto the corner boss as shown on the left of the second image).

I'll then have 2 15mm risers that connect to the other two corner bosses, and provide a chunk that's thicker than the bosses in an attempt to have the riser freeze after the boss.

Any thoughts? Madness?

 

Great work!!

This is my inexperienced take on your issues.

If you look at how the mold is feeding, what is the value of the risers? The metal will fill above the riser gates quickly and they will start to freeze. All of the air must be expelled through the sand anyway to fill the pattern which you already did successfully. I think the risers are just taking metal and making the mold fill slower and should be eliminated.

How did you calculate the gate size on the feeder? If you make it the thickness of the flange you'll eliminate the shrinkage issue. If you need more square inches just make it wider.

I would put a well under the feeder and a short runner in the drag to let the metal rise into the mold instead of splashing through the gate. And designs I've seen have a smaller bottom diameter of the sprue to choke the flow.

Somebody please correct my misconceptions.

 

I take your point re the risers. Maybe I'll keep the blocks at the bottom of the 15mm risers in the pic, but not bother cutting the risers to the top - that way there's a (small) thick bit to freeze after the corner bosses. Would I call this a bob?

And I'll poke wires through the fins - I don't mind wires as these will get a pass on the mill anyway.

 

If I understand you correctly, you are casting the part with the fins up (fins in the cope).

The defects don't really look like shrinkage to me. They look more like sand issues.
The reason I say this is because the defects on the side look almost exactly like the defects on the top.
The shrinkage issues I have had look like parallel horizontal tear lines and larger depressions, not pits like you have.

I cast a piece similar to this, perhaps slightly smaller, and I cast it fins-down (I had bosses on top instead of fins).
But I did not have any problems with my castings, either with sand defects, or shrinkage.
The only problem I had was a large bubble of air that got trapped in the top of my sodium silicate mold, due to me forgetting to vent the top of the mold.
Some say that sand does not need to be vented, but that only applies to sand that is porous enough to vent itself. Bound sand like sodium silicate and resin-bound definitely need a small vent hole at the top of each high point in the mold (in the cope).

I used dual runners, and a sprue with basin below it.
The runner is generally suppose to extend beyond the gate, as others have mentioned, to sweep loose sand down into the dead-end of the runner.
With bound sand, sometimes I dispense with the dead-end on the runner since the sand is bound very tightly.

The gate should choke the flow down, and be tapered generally from top to bottom, so as to scrape off any slag that is floating on top the stream of metal.
You need a steady fill, but not so fast as to cause turbulence, which can churn air and sand into the mold cavity.
The sprue needs to stay full at all times while you are pouring.

I put the runner in the drag, and put the gate at the top of the runner, so that the runner(s) fill first, then the gate does its scraping action.
You don't want the mold to begin to fill until the runners are full.

I have oversized my runners with the intent that they act sort of like a shrink bob.

And the number one cause of sand defects is pouring metal that is too hot.
I pour at 1350F, and measure the melt just before I pour with an accurate pyrometer.
I got a huge improvement in surface finish when I used the right pour temperature.

And I first tried to cast my part right side up, but I have always gotten a worse finish on the top of the cast part.
If you get a slightly worse finish on the inside of the pan, nobody will see that anyway.
My setup yielded an excellent finish inside and out.

And don't overheat the melt, since that has caused porosity in my pours (small pinholes all over).
Heat the metal as quickly as possible, but don't overheat, and don't overshoot 3150 F.

I have used both virgin 356 and the stuff I get from Art B, and have not seen any difference.
Any metal that I have overheated or metal that I have let sit for a while after it reaches pour temperature has had pinholes.
Pour immediately when you hit 3150 F.

I will post some pictures of the similar parts that I cast (I will add a link here in a minute when I find those photos).

Good luck.
Just my experiences with a similar piece.

Edit:
In engine work (steam engines in my example), the books say to keep the wall thickness of any part as uniform as possible, to reduce or eliminate shrinkage.
I have bosses on top of my part, and fins underneath, plus bosses at the bolting points, and no shrinkage, so there is some leeway in the thickness uniformity rule.
And I use a generous fillet at all corners. Sharp transitions/corners always seem to shrink/split.

Edit:
I put this info here:
(starts on post 66)
http://forums.thehomefoundry.org/in...-burner-bound-sand-molds.111/page-4#post-7617

Edit02:
I use as small a drill as I can find (a 12" long bit), and drill from the inside out to prevent breakout, in the highest points of the cope mold. With non-bound sand, you can just poke a find rod through with your hand.

 

It's real easy to poke from the top, and use a gauge so you stop 1/2" or so from the pattern. If you hit the pattern the little bit of spike is easy to remove.

Are the corner bosses thick, or is there a step inside? All changes in thickness contribute to shrinkage issues, of course. If they get drilled, can you put a hole through them when casting?

 

I did overshoot on the temp a bit... I measured and it was too low so I let it run for a while, then put the pyrometer back and it was ~850 C (1560 F). I immediately shut off the gas and skimmed the dross, then gingerly lifted the crucible out into the pouring shank. It was probably still too hot, and I didn't degas since the crucible was so full. This is with virgin LM4, which has a recommended pour temp of 790 C (1450 F)... no idea what kind of part that recommendation is for though!

I am indeed casting fins-up, as I didn't think I would be able to hang the mass of the bowl part of the sand upside-down. An as a result of making that decision, I made the drag very shallow (to save sand / mould weight) since there's no part in it - it's only 45mm high and there's no way it would survive being flipped even if it were flat. I took inspiration from this video of a rocker cover:



Admittedly they have rather nicer runners and gates! If I had more metal capacity I would've done something similar, but for the moment I'm sticking with the furnace I've got.

The corner bosses are thick, and get blind holes from the mating face in machining so I can't make a hole right through (the holes get oily). I could maybe put a dimple in the drag to make it a bit less heavy but I was worried about a dimple of sand breaking off when pouring, or the dimple ending up off-centre and making it hard to drill the hole in the correct location.

Thank you again for the help!
--
Olly

 

I enjoyed the video. Thought it was interesting that he missed the riddle every time he dumped sand into it.

 

I assume the procedure says "use the riddle". Boss, I used the riddle!

 

It is interesting that they used a number of risers (bobs?) on one side at the runner.
I have also noticed that automotive blocks cast in aluminum use a serious number of risers, which allow the main casting to solidify before the risers solidify (thus avoiding shrinkage).
I can only guess that my oversized runners are acting as risers/bobs, because I don't get any shrinkage on my engine base castings.

 

Look up a member on AA named demag. He knocks it out of the park with his valve covers.
http://www.alloyavenue.com/vb/showthread.php?5585-My-latest-casting

 

I have seen and admired demag's work on many occasions! Excellent results there.

Okay, bit more rework on the feeder. I have added a runner in the drag that has a dead end, and have tapered the gate:

 

I just saw demag thread off the post above. Quite nice! Looks like his only riser is off the top.

Like your gate modification.

 

There are a lot of different ways to do gating, and everyone has their favorite method.
I was reading about knife gates, and so I elongated my gate, not quite as long a a knife gate, but much longer than my typical gate, and made the entry height of the gate the same height as the flange on the part.

It worked well for me, so you may consider it.
If the gate is the same height as the flange, then it makes it much easier to cut the gate off and clean up that spot (photo of my similar mold attached).
My runners are deeper than they need to be, and I ended up cutting some of the runner off to make a taper towards the gate. Deep runners work fine if you have enough crucible capacity.

And generally the gate starts wide (one some commercial patterns I have seen (photo below), and tapers both in height and width, with the smallest part at the pattern.
I think the idea is to scrap off slag and impurities that float on top the melt.

Edit:
The last photo is a closeup of my elongated gate, which makes cleanup once the gate/runner is cut off easy.

 

Different metal (cast iron), but my green sand venting observation may still apply:
I started out venting molds with a coat hanger wire from inside my mold, but found it caused more poking out of sand than I liked and caused some shrink defects where the vents originated on the cast part. So, I switched to 1/16" spring wire (also known a piano wire) because it is stiffer and straighter than ordinary wire. I bent a tee handle on the one end to make it easy to hold in my hand and manipulate. I also ground a point on the end.

I like using the finer gauge wire much more than the larger wire. It seems to provide adequate venting since a lot of air can quickly run through a small hole and iron runs up the vent hole sometimes as much as 2 inches. No more shrink defects. And there is minimal poke out. I do place one hand on the outside of the mold to support the sand as I push the wire through from inside to further minimize poke out. FWIW.

Denis

 

Sounds like it can be done whichever way suits me best then! I've just found a paperclip so I'll sharpen that and see how I go.