Rider Ericsson Homemade Castings

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.
All good ideas and images. I have a plan now. Thank you all very much. 🙂 🙂 🙂

Please expound on the "retracts" and "cement and clamp the mould pieces" together. I think you are saying to cement the flasks together, and not the as cast pieces, right?

I love learning about casting things. I read a previous post made and used the reference you posted from the archives about core patterns and core prints.
 
Here is an example of retracts (retractable pieces) by the light pole manufacturer "Spring City".

Everyone who does hobby casting has to figure out how complex they want to get with their patterns and molding.
There is no "right" or "wrong" way in the hobby.
The "right way" is what gives you the cast part that you desire, with the quality you desire, using the methods and materials that you prefer.

Bound sand, such as with sodium silicate, opens up a lot of possibilities, and you can make entire molds from sodium-silicate bound molds, not just cores. SS molds can be form-fittted closely around a pattern, sort of like how a crankcase cover on a motorcycle closely covers the flywheel.
This saves a lot of sand and sodium silicate.
.


I started playing around with making twinned sodium silicate molds with a small pattern that was a cylinder support, and the intent was to avoid having to create two of the same patterns (photos attached).
Obviously you can make two separate molds, and pour them individually, but if you cement the molds together, and clamp them, you can use one pattern to make two molds, and pour two castings at the same time.

Retracts were used to make molds for large flywheels, where they would only make a pattern for one spoke, with a small pie-shaped section of rim and hub, and then they would use a retractable spoke section, and make the flywheel mold on the floor, one pie-shaped section at a time, thus saving a huge amound of pattern making effort, not to mention the space savings of only having to store a single spoke/hub/rim section instead of a huge flywheel pattern.

I only had one finished cylinder pattern on the green twin engine, and used it to cast two cylinders at a time. (photos attached)

A little creativity goes a long way in the casting world.

.
rImg_1869.jpg
rIMG_8917.jpg
rIMG_8915.jpg



rImg_1672.jpg


rImg_1673.jpg


rImg_1676.jpg


rImg_1681.jpg
 
Last edited:
I simply did not draw them as it was not needed for the example. I would have thought it was obvious they they would pull out the sand OK with the position of the split line as mentioned. I also did not show draft

split.JPG


But as Crest says they will all cast with the method we show and no risk of a foundry loosing the loose piece (retract). Yes it happens and I have made replcements for people where the foundry has lost them.

I don't really see the advantage of sticking those two cylinder moulds together unless you only have one flask. At least it is in a position where alignment of the joint does not matter, you would have to get it right if this yoke were cast as two moulds to avoid a step in the finished part
 
Last edited:
here is one half with the bosses added

rider yoke pattern half.jpg


And the mould that can be taken from it. Just drop a simple core into the space left by the core prints on the pattern and you are good to go

rider yoke mould.jpg
 
Yes, I see it now.
Very clever setup.

For my cylinder, yes, only one snap flask, and only one cylinder pattern too.

And so I assume crest set his mold up about the same way.

I like your arrangement much better than my cemented molds (for this part), but cemented molds do have their uses.

And I prefer a pattern without retracts, but sometimes those are also necessary.
I would screw the loose part(s) to the main pattern when the pattern is being stored.

The wishbone pattern could also be made in one piece, and a crude follower formed up to allow the two mold halves to be made.
I have noticed that most patterns I have seen at commercial foundries (old and new foundries) are generally one-piece, with the exception of something like matchplates.

All of the Cretors original patterns I saw where one-piece.

So I guess I was thinking a one-piece approach.
Small patterns with thin parts like spoke halves can be fragile, and I have gotten into the habit of making mostly one-piece patterns.
I actually shattered one of the green twin flywheel pattern halves, but was able to glue it back together.
That pretty much cured me of two-piece patterns, but in many cases two piece patterns are not optional, but instead a necessity.
.

Edit:
For 3D printed patterns, often you get forced into printing pattern halves, just to have a flat side to start the print with.

.
 
Last edited:
All of the wooden patterns are completed and I got a free weekend to start casting! Hot and dry weather here in central Florida. The first part I wanted to attempt is the leg so I could practice with sand consistency and the parting line. Unfortunately I ran short of sand using my larger flasks so I had to switch to smaller flaks and smaller parts. Ordered another 24 lbs of foundry sand.

Flywheel Bracket
this split pattern was made with melamine board and small bits for the webs glued on and filleted with milliputty. Numerous coats of enamel paint, sanding and filling before they passed inspection. Then a coat of paste wax and buffing.

flywheel bracket patterns.jpg


I recently purchased a generic paint/grout mixer for a tile job I did for one of my kid's house. Used a metal bucket and mulled up the sand using it. Can only do about 5 inches worth or it bogs down the motor, but it worked decently for the $50 I spent on it. (It really works great for tile adhesive.)

mulling equipment.jpg


The forge is a 4 year old propane fired and ceramic fiber insulated one made out of an air tank. It melts up enough metal in about 25 minutes. My temperature gun didn't seem to be working today and the first part was a little cool. The casting sand was also a little dry and one can see where I lost the edges in two places. I rammed up and poured another while I had everything setup and It looks better. I gated it into a flat machined area instead of the pivot point. I haven't cast anything in a little over a year so I'm happy with these parts and will move on to some other small flask parts tomorrow.

forge equipment.jpg
small flask.jpg
two attempts.jpg
second attempt.jpg
 
Nice work but I would avoid feeding directly into the part and use a runner instead.
If there is one point you get problems it is at the sprue base where turbulence is greatest.
Rising from the part is ok.
Rich
 
Thank you Rich,
For the previous post:
Just so I get it right.....The part on the right was fed directly from the sprue and the part on the left had a runner?


What about making two parts in one casting? I have seen others make small parts on "trees"
Would that work on two different geometry parts? The image below is what I was thinking...
two parts.png
 
You can certainly cast more than one part in a flask at a time. Quite a few commercial foundries will charge by the flask so you may as well get your monies worth. I've just finished building an RLE engine and this is a photo of how they cast some of the iron parts using the separate patterns and squeezing as many in as they could. If you only have a limited number of flasks then it can be a more economical way to cast as you don't have to keep reheating to pour each part individually.

rlepat.jpg


This is similar to using match boards where several parts can be mounted onto the board and the runners included so you don't have to keep cutting the runners into the sand but not really needed for one off hole casting.
 
Here is another option that may work, in blue, and rotate the pattern on the left 90 degrees.

Use a short sprue, no pour basin, smooth transition into a runner that has an offset spin trap at the end that is vented out the top of the cope, and knife gates into the parts.

Nobody seems to use knife gates regardless of the part, but they are very useful when the gate enters a thin section.

Hobby folks almost universally use a pour basin, but I don't know of any better way to churn air, slag, and bifilms into aluminum than to use a pour basin. A pour basin is not required, and I never use them.
A tall spure also churns a lot of air and trash into the melt due to the high velocity it creates.

Sometimes I use a 1" tall section of 3" pipe laying on top the cope at the sprue, as a sort of catch wall.
You can practice pour technique using water, but don't put water into a crucible unless you bake it dry afterwards.
Pouring is non-linear; you fill the spure as fast as possible, and then reduce the flow when the runner is full, but keep the sprue full at all times.

High velocity is your #1 enemy.
You want a somewhat slow, smooth, unbroken flow of metal into the mold cavities.
And I use a 1/16" vent at the high point of each mold cavity, to relieve any trapped air.
Some folks don't vent, but better to vent and not need it than to not vent and potentially trap air in the mold as has happened to me.

Abrupt changes in the path of flowing molten aluminum breaks the bifilm exterior of the flowing metal, and churns slag, air and bifilms into the mix, and can also dig into the sand mold and cause sand inclusions.
Enemy #2 is abrubt changes in the flow of metal.

Having a spin trap at the end of the runner keeps the metal from hitting a dead end, which causes a pressure wave back through the melt, and splashes metal into the mold. You want to avoid splashing anywhere.

You may need a riser in the center of the runner between the two patterns.

Gates at the top of the runner, so that they scrape off slag that normally floats on top the melt.
Never let the sprue go dry once you start filling, otherwise you will aspirate air.

The flow rate is controlled by the gates.
Don't try and control flow rate (velocity) by the sprue base.
.

48Image1.jpg
 
Last edited:
Hi Raveney,
Just so I get it right.....The part on the right was fed directly from the sprue and the part on the left had a runner?
Yes you correct.
Jason and Pat have answered your other questions quite adequately
Rich
 
Here is another option that may work, in blue, and rotate the pattern on the left 90 degrees.

Use a short sprue, no pour basin, smooth transition into a runner that has an offset spin trap at the end that is vented out the top of the cope, and knife gates into the parts.

Nobody seems to use knife gates regardless of the part, but they are very useful when the gate enters a thin section.

Hobby folks almost universally use a pour basin, but I don't know of any better way to churn air, slag, and bifilms into aluminum than to use a pour basin. A pour basin is not required, and I never use them.
A tall spure also churns a lot of air and trash into the melt due to the high velocity it creates.

Sometimes I use a 1" tall section of 3" pipe laying on top the cope at the sprue, as a sort of catch wall.
You can practice pour technique using water, but don't put water into a crucible unless you bake it dry afterwards.

High velocity is your enemy.
You want a somewhat slow, smooth, unbroken flow of metal into the mold cavities.
And I use a 1/16" vent at the high point of each mold cavity, to relieve any trapped air.
Some folks don't vent, but better to vent and not need it than to not vent and potentially trap air in the mold as has happened to me.

You may need a riser in the center of the runner between the two patterns.

Gates at the top of the runner.
Never let the sprue go dry once you start filling, otherwise you will aspirate air.

.

View attachment 156751
Pat,
I disagree and so would olfoundryman who uses them all the time. He keeps the sprue choked to control flow and reduce air ingress.
I have to admit I have broken the rules at times in order to achieve a successful casting but most of the time I use a pouring basin with a slight raised lip at the sprue entry.
Rich
 
It is well documented (I think) that pouring into the curved-bottom basin creates metal that rotates back upwards and then collapses back on itself.
You can clearly see it in the simulations, and this is the worst thing you can do to aluminum, since it entrains air, bifilms, and slag into the melt.

Everyone does it their own way, and everyone has their favorite arrangement.

I can only state how I do it, and why I do it.

For aluminum, you can get away with a lot, but I have seen sectioned castings, and you can see bifilm defects and slag inclusions when there is a lot of turbulence. Outwardly, the casting looks very good.
For structural parts, you really need to eliminate the bifilm and any other defects.

For hobby parts, often anything works ok.
I have seen a lot of aluminum hobby parts that could be improved upon, and I know why they are getting surface finish and other defects.

To each their own, but these are my recommendations.

Edit:
Olefoundryman is one of the best aluminum casting persons I have ever seen, but the best I have seen is Bob Puhakka, and I have xray photos of his castings, and they are indeed flawless inside and out, without bifilm defects.
Bob Puhakka uses modified John Campbell methods, and so do I, although small castings required slightly different methods than the big castings that Bob makes.

When molten aluminum is flowing in a pour basin, down the spure, down the runner, etc. it is creating a thin solid film where it touches the sand mold. If the velocity is slow, then this film remains on the sides of the metal flow, and does not go into the mold cavity.
If you splash into a pour basin, use irreguarlar passages, high velocity, etc., then the film is broken, and the molten metal rolls over the bifilm, creating a defect because it is a cold joint.
Bifilm defects can be small or very large and noticeable, but they should be avoided if possible, and must be avoided for commercial structural parts.

.
 
Last edited:
Hmmm.
Well watching Bob's video on bifilms does indicate how a pouring basin can aid the fill without bifilms being drawn in.
 

Attachments

  • Puhakka.jpg
    Puhakka.jpg
    103.2 KB
Thank you for the tips and discussions. When I first started sand casting I used the tin can rammed up with sand as the pouring sprue as I observed lots of people doing that. I switched over to small pouring basins cut into the top of the cope and use a SS tapered sprue for these pours. The basin has sharp edges transitioning to the walls so the metal doesn't ride up the wall as Pat mentions. The film appears to stay unbroken using this technique. I watched an Oldtimer day before yesterday and he vehemently cussed the blokes that used bean cans as I used to. Sounded like he knew his business.


The pours I did today were done immediately after getting the ingots to melt, and only giving enough superheat based on how long it took to take a sip of gatorade, put on my leather apron, leather sleeves, gloves and the face shield. Today's bounty all came out very good. The biggest improvement I think is that I am using real foundrymen's parting powder instead of baby powder. Every pattern pulled cleanly!

Displacer Yoke
This pattern was made as many suggested here and I made the core using just sand rammed up in a pool chemical bottle split and taped back together with painters tape.

displacer yoke pattern.jpg

pattern n core.jpg


cope.jpg
as cast.jpg


No silicone CO2 or Molasses on this simple shape was required. The casting has some flashing especially where the core meets the pattern, but I cleaned it up on the beltsander without much effort.

part 02A.jpg
part 02.jpg
 
Last edited:
Good to see the way Bill & I suggested worked out for you, when do you start on the bigger bits ;)
 
Thank you for the compliments. I am very appreciative of the interesting build logs and innovative ways that people use to create things. 😁The core print for the yoke was something that never occurred to me.

Walking Beam and Support

I used 3mm board and 1/4" plywood along with tapered ends of dowels to glue up the patterns while the two halves were fastened together with double sided tape. Fillets were made with milliput again, but I tried using matte clear shellac to reduce the fibers swelling before painting with enamel spray-paint. The support is a composite of 14 individual pieces of basswood glued up. Some round endmill work was required to get the contours accurate enough.

Once again I was pleased with how cleanly the patterns pulled using the professional parting powder. The beam looks very very good, and the support has some shrinkage on one side that I will fill in after machining the rest. Perhaps the gate wasn't big enough, or I should have added a reservoir at the base of the sprue to help with that?

20240505_103109.jpg
20240601_110845.jpg
20240601_131311.jpg
20240601_132106.jpg
20240601_132116.jpg
 
It is well documented (I think) that pouring into the curved-bottom basin creates metal that rotates back upwards and then collapses back on itself.
You can clearly see it in the simulations, and this is the worst thing you can do to aluminum, since it entrains air, bifilms, and slag into the melt.

Everyone does it their own way, and everyone has their favorite arrangement.

I can only state how I do it, and why I do it.

For aluminum, you can get away with a lot, but I have seen sectioned castings, and you can see bifilm defects and slag inclusions when there is a lot of turbulence. Outwardly, the casting looks very good.
For structural parts, you really need to eliminate the bifilm and any other defects.

For hobby parts, often anything works ok.
I have seen a lot of aluminum hobby parts that could be improved upon, and I know why they are getting surface finish and other defects.

To each their own, but these are my recommendations.

Edit:
Olefoundryman is one of the best aluminum casting persons I have ever seen, but the best I have seen is Bob Puhakka, and I have xray photos of his castings, and they are indeed flawless inside and out, without bifilm defects.
Bob Puhakka uses modified John Campbell methods, and so do I, although small castings required slightly different methods than the big castings that Bob makes.

When molten aluminum is flowing in a pour basin, down the spure, down the runner, etc. it is creating a thin solid film where it touches the sand mold. If the velocity is slow, then this film remains on the sides of the metal flow, and does not go into the mold cavity.
If you splash into a pour basin, use irreguarlar passages, high velocity, etc., then the film is broken, and the molten metal rolls over the bifilm, creating a defect because it is a cold joint.
Bifilm defects can be small or very large and noticeable, but they should be avoided if possible, and must be avoided for commercial structural parts.

.
Can you explain "bifilm"?
 

Latest posts

Back
Top