Here is a rough example of the spin trap that I use.
The sprue should be kept short, not much above the top of the drag.
The sprue should transition smoothly into the runner (don't use a rough transition like the screencap below).
The gates should be on the top of the runner.
I generally put the top of the runner at the top of the drag, and keep the gates in the drag (you can flush the gates down into the top of the runner).
The runner enters the spin trap an a tangent, again with a smooth transition.
The spin trap goes out the top of the cope.
I drew it at a random height, but it stops at the top of the cope.
The lip of the crucible should be as close to the top of the sprue as possible when the pour starts.
The pour should almost immediately fill up the sprue, and then the flow must be continuous and never interrupted during the mold fill.
The initial metal/air mix flows down the V-shaped runner, sweeping the sprue and runner of any loose sand.
The spin trap swirls the metal/air mixture and prevents any bounce-back on the runner.
A bounce back in pressure on the runner would shoot a stream of metal through the gates, which you don't want.
As the metal reaches the top of the runner, you have established a very hot stream, and have heated the entire runner before mold fill begins.
The gates act to skim off any floating slag on the metal.
The gates should be configured to avoid waterfall, and there should be enough gates to fill the mold quickly but without turbulence.
I often use a U-shaped runner, with gates on either side of the mold cavity, to fill from both sides at once.
There would be a spin-trap at the end of each runner.
For thin castings, I use one or more knife gates, to keep the velocity down and get a complete fill.
For thicker castings, I use a gate shaped more like the one shown below.
I try to gate into the thickest part of a casting, and also gate into a part of the casting that has to be machined off, to erase the gate.
I use risers if there are thick parts on a casting, to avoid shrinkage.
If possible, the casting thickness should not vary, but sometimes this is not possible.
For castings with varying thickness, you have to avoid the thinner parts solidifying first, and drawing from the thicker parts (by using appropriately sized and spaced risers).
The runner and spin trap ensure that the cooler contaminated metal goes to the spin trap, and not into the mold.
If you section a spin trap, you may see entrained air bubbles, sand, and slag.
I don't run calcs for anything, since I am not having casting defect problems.
Using spin traps requires a bit more molten metal, and so maybe a larger crucible.
I always want a casting to turn out well on the first pour.
I despise having to pour something more than once, since my sand is not reusable, and so far I have not had to do that with iron in recent years.
The photo below shows what would be on one or both sides of the mold cavity.
I sometimes place the entire pattern in the cope, and fill upwards.
I vent the highpoints of the cope mold to avoid trapping air in the mold (I have had air get trapped and ruin a casting), using small holes, perhaps 1/16".
The sprue should be kept short, not much above the top of the drag.
The sprue should transition smoothly into the runner (don't use a rough transition like the screencap below).
The gates should be on the top of the runner.
I generally put the top of the runner at the top of the drag, and keep the gates in the drag (you can flush the gates down into the top of the runner).
The runner enters the spin trap an a tangent, again with a smooth transition.
The spin trap goes out the top of the cope.
I drew it at a random height, but it stops at the top of the cope.
The lip of the crucible should be as close to the top of the sprue as possible when the pour starts.
The pour should almost immediately fill up the sprue, and then the flow must be continuous and never interrupted during the mold fill.
The initial metal/air mix flows down the V-shaped runner, sweeping the sprue and runner of any loose sand.
The spin trap swirls the metal/air mixture and prevents any bounce-back on the runner.
A bounce back in pressure on the runner would shoot a stream of metal through the gates, which you don't want.
As the metal reaches the top of the runner, you have established a very hot stream, and have heated the entire runner before mold fill begins.
The gates act to skim off any floating slag on the metal.
The gates should be configured to avoid waterfall, and there should be enough gates to fill the mold quickly but without turbulence.
I often use a U-shaped runner, with gates on either side of the mold cavity, to fill from both sides at once.
There would be a spin-trap at the end of each runner.
For thin castings, I use one or more knife gates, to keep the velocity down and get a complete fill.
For thicker castings, I use a gate shaped more like the one shown below.
I try to gate into the thickest part of a casting, and also gate into a part of the casting that has to be machined off, to erase the gate.
I use risers if there are thick parts on a casting, to avoid shrinkage.
If possible, the casting thickness should not vary, but sometimes this is not possible.
For castings with varying thickness, you have to avoid the thinner parts solidifying first, and drawing from the thicker parts (by using appropriately sized and spaced risers).
The runner and spin trap ensure that the cooler contaminated metal goes to the spin trap, and not into the mold.
If you section a spin trap, you may see entrained air bubbles, sand, and slag.
I don't run calcs for anything, since I am not having casting defect problems.
Using spin traps requires a bit more molten metal, and so maybe a larger crucible.
I always want a casting to turn out well on the first pour.
I despise having to pour something more than once, since my sand is not reusable, and so far I have not had to do that with iron in recent years.
The photo below shows what would be on one or both sides of the mold cavity.
I sometimes place the entire pattern in the cope, and fill upwards.
I vent the highpoints of the cope mold to avoid trapping air in the mold (I have had air get trapped and ruin a casting), using small holes, perhaps 1/16".
