Sand Casting Design Rules

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Well-Known Member
Staff member
HMEM Supporting Member
Global Moderator
Jul 2, 2021
Reaction score
MidSouth, USA
On pages 7 and 8, you can see an offset pour basin.

I think offset pour basins are a very bad idea, and they basically churn the metal, and entrain air and sand into it.

I never use a pour basin, but instead pour directly down the tapered sprue.
I use the gates to control velocity into the mold cavity.
I don't try to use the sprue or runner system for velocity control.

Page 7, I do use a runner extension, and I use a spin trap at the end of each runner, which is just a round hole similar to the sprue, but not tapered, that is offset from the centerline of the runner, so the metal enters the trap and spins, to prevent a sudden impact on the runner pressure.
The spin trap is open to the top of the mold.

I use gates on the top of the runner, and generally try to put the gate at the bottom of the mold cavity, so that the mold cavity fills upwards, if possible.

Page 10, I don't use a well at the bottom of the sprue, since this churns air, slag and sand into the molten metal flow.
I use a smooth radius transition from the bottom of the sprue into the runner(s).

Page 11, I don't step the runner(s), ie: I use the same runner section for the entire runner.

Page 12, I try to have the gate(s) enter the mold cavity at the bottom of the cavity, and fill upwards.
You want to avoid the gate(s) at the top of the mold cavity, since this will give a waterfall effect, and entrain air, sand and slag into the casting.

I never use bottom gating (ie: a horn gate).
I consider this an old and obsolete method.

The intent of my sprue/runner/gate/spin trap system is as follows:

1. Pour down the spure as fast as possible, with the lip of the crucible as near the sprue opening as is possible.
Don't oversize the sprue, and use a tapered sprue.
The sprue should fill in a second or two, and you should keep the sprue completely full for the entire pour.

2. The molten metal flows down the runner first, and any entrained air, slag, loose sand, etc. flows into the spin trap, where it will remain.
Slag floats on top of the molten metal flow generally, and so the gates on top of the runners help scrape off any remaining slag in the molten metal flow.

3. I use generous runners, with the idea that the initial flow of metal will be cooled in the sprue and runner(s), and so the flow of metal into the spin trap not only sweeps the runner system clean, but also heats the runner system, so that only very hot molten metal enters the mold cavity via the top-mounted gates, once the runner is completely full.
The gates control the velocity of the molten metal into the mold cavity.
You want to fill the mold cavity as fast as possible, but at the same time you want to control the velocity of the metal stream so that you do not get splashing inside the mold cavity.
You want a smooth fast laminar fill of the mold cavity.

4. I use risers when there is a variance in the casting sections.
Sometimes I have gotten away with one central riser at the center of a flywheel mold, and other times I have needed a few risers spread evenly around the rim of a flywheel mold.
I leave the risers open to the top of the mold.

5. Ideally the gate(s) should enter a section of the casting that has to be machined, so that the gate entry is completely removed.
This saves some finishing work.

I do not use a filter with this sprue/runner/gate/spin trap layout.

Last edited:
I ran across this paper by Santosh Reddy Sama and Guha P Manogharan, and it contains some good foundry information.

I will also note a few things I have learned from a professional foundry owner, and compare these things with what is described in this paper.

Use large radius where can it pull mold and and less shiny problems.

Gating is a eyeball by most pattern makers. But making thousands of parts you can adjust. Pattern also helps too it like wheel hub make the tapper hold center will help shrinkage.

Hi Pat
I have never done any casting but find it very interesting and read most of the home foundry posts.
In your post above you have used a few terms that confuse me and if you do not mind clearing my confusion it would be appreciated. :)
Many times I hear the terms gates and runners used interchangeably and you use them as two different things. Would you tell me the difference? My uneducated guess would be ( in your example above) The runners on the bottom feed the different cavities and the gates at the top would equalize pressure between them ??? And the gate would vent to the top ??

Thanks in advance


Looking at the screencap, which is from the pdf listed in the first post, we have the pour basin at the top of the cope (cope = top half of the sand mold), then the tapered sprue, which is where the molten metal travels vertically into the sand mold.

Then there is one or more runners, which carries the molten metal from the bottom of the sprue horizontally two a location that is close to the mold cavity, but outside the mold cavity.

And finally the gate(s) connect the runner(s) to the mold cavity, and the mold cavity is filled through the gates.

A riser is a void in the sand (generally in the cope) that contains an amount of molten metal over an area of the casting that is abnormally large compared to other parts of the casting.

Risers are required when a casting has unequal parts, such as a large section area, and also a small section area.
The small section will solidify first, and shrink as it solidifies, and so to prevent the larger section from feeding molten metal into the thinner section while the thinner section is solidifying, a riser is used above the larger section, so that metal can be drawn from the riser as the casting is cooling and contracting.

A lack of a riser can cause shrinkage defects, and often times hot tears, where parts of the casting tear apart while the casting is in the semi-molten state, due to shrinkage.

The article in post #1 mentions a "gating system", but I don't think that is a good term.
I can't recall a better definition of the molten metal distribution system, but that is what it is, a distribution system that consists of sprue, runner(s), gates, and risers.

There are several ways to control the molten metal velocity, which are to constrain flow at the sprue base, by using a filter, or by controlling flow via the gates.
I use a gate-constrained velocity system, with no filter.

I also use a spin trap at the end of each runner, which gives the loose sand, slag and entrained air a place to escape.
The spin trap is offset from the centerline of the runner, so that the molten metal swirls in the trap, with the intent to avoid a sudden pressure shock in the runner which would occur with a blind dead-ended runner.
The spin trap vents out the top of the cope.

Any sudden pressure changes in the molten metal cause it to discharge metal through the gates into the mold cavity in a way that splashes and entrains a lot of air, as well as breaking up the molten metal into separate parts/pieces.

A smooth, even laminar flow is desired when filling a mold cavity, with low velocity, no turbulence, and no splashing.


Last edited:
Here is an very good example of how not to fill a mold cavity.

A high velocity molten metal stream that is introduced in the wrong area of the mold cavity causes the metal to form a wave, just like in the ocean, and when the wave rolls over, it entrains air, slag, sand, etc into the metal.

Many model engine castings suffer from defects cause by high velocity mold fills, and by a poorly designed runner and gate system.

And cooling a gray iron casting quickly can cause the dreaded "chills", which are spots that are approximately the hardness of tool steel.

Generally a small amount of ferrosilicon is added to iron melts to prevent thin sections of the casting from developing chills.

There is absolutely no reason to have any defects in a gray iron casting.
One does need to to pay close attention to the fundamental dynamics of mold filling in order to make defect-free castings.

Last edited:
Here is a better method of mold filling.

Not quite a laminar flow, but not too bad.

The velocity is kept low, and sufficient gates are used (often wide gates on a thin part) to fill the mold cavity relatively quickly without excessive turbulence.

Generally you want to fill from the bottom upwards to the top, or from one side to the other, so that you don't get a cold join where to wave fronts meet.

Thanks Pat ! that explains it very well.
I think what threw me was your statement in post number 2 " I never use bottom gating"
Which I took as gates at the bottom of the mold and not at the bottom of the runner.

Again, thanks for the great explanation ! The confusion has cleared :)

I had to go back and re-read post #2 to remember what I was talking about.

I did not make the "bottom gating" reference clear at all, but what I was referring to was a horn gate, or something like a horn gate.

You are correct that I always put the gates at the top of the runner(s).

Some illustrations of horn gating.


I have chatted with one individual who runs his own aluminum foundry in Canada, and he uses John Campbell's methods.

There is a lot of debate about John Campbell's casting methods; what they mean; how effective they are; are they correct; etc., etc.
Everyone has an opinion about John Campbell and his 10 Rules for Good Castings (he has a book, which contains the same 10 rules).

The 10 rules can be found on this website:
The guy in Canada says that most foundries these days set up their sprue/runners/gates/risers wrong.
There is also much debate about the guy from Canada and his statements.

Bottom line is the rejection rate of any foundry's castings.

If a foundry has a high rejection rate, then it is not doing its sprue/runners/gates/risers correctly, regardless of who or what you believe.

And if a foundry can xray its castings and prove that they have zero internal defects, and low or no rejection rate, then that is the methods that may be wise to study and attempt to mimic on a hobby scale, which is what I do.

One of my mottos is "Talk is cheap; show me your castings".
Anyone should be able to section their castings to look inside for defects, and machine the castings to test for chills/hard spots.
There should not be any internal or external defects, and no hard spots in the castings.

If you find defects in your castings, you are doing something wrong.


Latest posts