Adventures in Sand Casting

[Aiden CC]

I don’t like casting, and it doesn’t like me. That being said, every once in a while it’s useful to be able to do it, and this is one of those times. I want to make bronze rivet bosses for sax sheaths. I made the master below out of three pieces of brass, and with 4-6 per sheath I would rather not have to take the time to shape each one!

 

Not shown is a long  0.093” shank that I’m holding in the picture. This is the geometry that I want to replicate, and here is how my first attempt came out:

 

This is actually a lot better than I thought it would turn out, but still leaves some to be desired! With how many of these I want to make, it would be nice to have minimal cleanup. 
 

A few questions:

1) Is it a fair assessment that the the larger defects in the above casting come from loose/missing sand? I disturbed the sand a bit when making the mold unfortunately and may have dropped some in through the the spruce. 
2) Is a 10 weight pct tin bronze a good alloy for this type of casting? I want to do a Cu/Sn bronze, but am somewhat flexible there. 
3) Will I be able to make this part acceptably with sand casting? It has fairly fine detail and I want an even surface finish. If not, I’ll make a silicone mold to pour wax into for investment casting (another can of worms!)
4) What is a good way to clean up castings like this? I don’t need a perfect surface, just a bit more even. 
5) Is there anything from my first casting that jumps out as something to do differently next time?

 

Thank you for any insight here, I’m definitely outside of my comfort zone with this experiment!

 

Edit to add one more question:

Would I be better off casting two parts for this, i.e. a washer and a rivet with a head and shank?

Edited November 21, 2022 by Aiden CC

[Jerrod Miller]

My first ever experience with casting was an awful failure.  You can certainly get there.  

 

Are you doing green sand or oil?  Getting the material ratios is critical for optimal surface finish.  This includes sand grain size ratios.  You will also want a much smoother gating system.  I'll try to return to this thread later when I have more time for more pointers.  

[Aiden CC]

Thank you @Jerrod Miller! I'm using a pre-mixed oil bonded sand. Would it be better to mix my own to get good control over grain size ratio? For gating, would it work to put in a larger sprue and then cut the gates directly into the sand? Ideally, I want to be able to clean up the casting by only removing material from the "back" of the rivet head, which seems like it would work with a gate cut into the sand on the back half. 

 

I think a longer term solution for these may involve a two part casting, it seems like maybe originals were done this way? It would certainly make it easier to clean up. Here is a picture I found which shows the general form I'm going for:

 

 

Some or all of these may also be hollow, but that's another can of worms. The buttons on the top sheath may be tinned, which could be another way to resurface them. I ordered some carving wax since I think I'll be able to create the fluting I want more accurately with carving than with filing.

[Jerrod Miller]

Other than getting your sand just right, here are a few things that I would work on for the mold/molding process.  

 

• I would cut the shank fairly short.  Just long enough to grab it to pull the head out of the sand.  This will still an impression in the cope sand that then gets drilled out, all the way out of the mold.  Cut the final pin length as needed.  
• You want the gating system much smoother in terms of metal flow.  This is especially important with copper base as it really wants to make air bubbles (Al too). 
  • Below is a very quick and dirty diagram (nothing calculated).  
  • Key features: 
    • Well at the base of the tapered sprue (with integrated funnel) to smooth out flow and transition from free fall to horizontal flow.  Ideally this would be parabolically tapered, not straight; but that is just overkill.  
    • Large opening to allow for faster pouring.  Fill up the sprue as quickly as possible.  
    • Runner the extends beyond the ingate; it is in the drag.  This is there to help flush oxides from the start of the pour and any sand that made it into the gating system to a sacrificial place, not the casting.  
    • Ingate in the cope, angled backward to the runner and tangent to the part.  This will help direct the metal in the smoothest way into the part.  
    • Not show in the image is that your stud/shaft should be drilled out to the open air.  This will act as a vent to let all the air out of the mold cavity as it is being filled with metal.  

[Jerrod Miller]

2 minutes ago, Aiden CC said:

Would it be better to mix my own to get good control over grain size ratio?

I would fix everything else design-wise first.  Then, if you still aren't getting the surface finish you want you can explore that.  

 

I meant to mention that you can totally hand cut the gating system into the sand, but it will be better if you can ram up the sand around a solid pattern.  Cut sand surfaces are not likely to be very smooth and the sand more likely to erode out into the casting.  And the sprue should be a bit taller than you want the final pin length to be.  This is so you definitely have a clean surface where you cut it to length, but also to give you head pressure (cupro-static pressure, in this case) in the mold.  

[Aiden CC]

Thank you for the very detailed advice! I'll try those changes out on my next attempt. I'll make an actual sprue pin instead of using a straight rod and try out using some wax strip to make the gating system. In my first attempt, the shank for the vent did go all the way to the top of the cope, though it only filled about halfway. I only really need about ~8 mm anyways though.

[Jeroen Zuiderwijk]

14 hours ago, Aiden CC said:

Thank you @Jerrod Miller! I'm using a pre-mixed oil bonded sand. Would it be better to mix my own to get good control over grain size ratio? For gating, would it work to put in a larger sprue and then cut the gates directly into the sand? Ideally, I want to be able to clean up the casting by only removing material from the "back" of the rivet head, which seems like it would work with a gate cut into the sand on the back half. 

 

I think a longer term solution for these may involve a two part casting, it seems like maybe originals were done this way? It would certainly make it easier to clean up. Here is a picture I found which shows the general form I'm going for:

 

 

Some or all of these may also be hollow, but that's another can of worms. The buttons on the top sheath may be tinned, which could be another way to resurface them. I ordered some carving wax since I think I'll be able to create the fluting I want more accurately with carving than with filing.

I believe they were shaped after casting. So cast a pin with a small thick disc, and then thin down the head and shape the head accordingly, probably using a die to create the correct head shape. 

[Aiden CC]

Thanks, @Jeroen Zuiderwijk that’s interesting. Do you think that the boss and rim on these are the same piece? It seems like the transition between them is quite sharp, but it’s hard to tell in any of the photos I’ve seen and I’ve only managed to find drawings of the version of these that has a wide, flat, head with carvings on top, showing that they are hollow. I sketched up a couple of general processes, do any of these look like what you were thinking?

 

[Jeroen Zuiderwijk]

Yeah a bit like that, f.e. like 3. I'm not fully certain, as I can't find any good images of loose examples at the moment. But one thing I've learned is that metalwork from that period is a lot thinner in general then you expect. If they could make anything hollow or thin walled then they did it to use the material as efficiently as possible. If I recall correctly, with type of rivets with the three holes in them, there are examples with the edges folded down, making them look thicker then they really are. 

[Aiden CC]

It makes sense that they would want to make parts like these hollow, especially since a casting needs some cleanup anyways. I was able to find a more zoomed in photo of those sheaths from above, and it does look to me like the boss and rim are different components on at least on the middle sheath, due to the shadows you can see in the flutes where they meet the central boss. Maybe a hollow, flat, head for the rivet and a thin sheet washer fluted with a punch and die or repousse? I think that the buttons on at least the bottom sheath could be formed from a single piece as they don't seem to have that same super sharp transition between the boss and the rim. Since I am still very new to casting, I think I'll probably start with thicker sections for the fittings I use to make it easier on myself while experimenting with other methods on the side. I'll try out casting the rivet and washer as separate pieces as well.

[Alan Longmire]

That looks like beaded wire soldered to a hollow boss.  Very frustrating to make at that scale, in my limited experience.  Could be made by making a thin washer and using repousse and chasing to create the beaded look.  Definitely two pieces.  

 

I have some articles on Anglo-Saxon jewelry techniques that may help.  Let me know if you want them, they're PDFs rather than links.  

[Aiden CC]

35 minutes ago, Alan Longmire said:

I have some articles on Anglo-Saxon jewelry techniques that may help.  Let me know if you want them, they're PDFs rather than links.  

That would be great, thank you!

 

I may try looking around at different types of pattern wire to potential make a master for casting these then. I think I've seen beaded pattern wire at various jewelry supply places. I do like soft metalwork like this, but at 4-8 of these on a sheath, that would be a lot of work for each knife, which may have been the point back then! I guess a hollow tipped punch could potentially do it more quickly as well.

[Alan Longmire]

PM sent. 

[Alan Longmire]

After looking at what I have used in the past, in addition to those articles (all public domain downloads, btw) I also used Theophilus (which I assume you have), Anglo-Saxon Crafts by Leahy https://www.amazon.com/Anglo-Saxon-Crafts-Revealing-History-Paperback/dp/0752429043,

 

Metal Techniques for Craftsmen by Untracht https://www.amazon.com/Metal-Techniques-Craftsmen-Decorating-Illustrations/dp/0385030274/ref=sr_1_2?crid=2I68Q347A1TK9&keywords=oppi+untracht&qid=1669145895&s=books&sprefix=oppi%2Cstripbooks%2C123&sr=1-2 ,

 

and The Complete Metalsmith by McCreight https://www.amazon.com/Complete-Metalsmith-Professional-Tim-McCreight/dp/1929565054/ref=d_pd_sbs_sccl_1_2/144-3537602-0534705?pd_rd_w=j4x5T&content-id=amzn1.sym.38bbd1de-73a5-4ef9-9954-df27c3112829&pf_rd_p=38bbd1de-73a5-4ef9-9954-df27c3112829&pf_rd_r=79T0Y2930JJ6N4S37J4M&pd_rd_wg=846yG&pd_rd_r=dda0a258-ab25-492e-a82f-dac47779a33a&pd_rd_i=1929565054&psc=1. 

 

I was lucky enough to find the Untracht book at a library book sale for $1.

 

As if you need anything else to read during grad school... 

[Aiden CC]

Thanks Alan, I’ll take a look at those resources! 
 

5 hours ago, Alan Longmire said:

As if you need anything else to read during grad school... 

Admittedly, it’s nice to have the occasional break from articles about the edge ductility of advanced high strength steels 

 

I followed the advice from @Jerrod Miller and it definitely improved things!

 

Here is a first shot at doing one rivet, the metal seemed to flow well even in a pretty shallow channel. 
 

I managed to get two at a time to work as well. I had some issues with sand erosion ending up in the castings, and, to be fair, I was warned. Next attempt will use solid forms for the runner and gates. I will probably try to do four of these in one flask, to actually save effort this way, I think I need at least two at a time, but four would be great. 

Edited November 23, 2022 by Aiden CC

[Jerrod Miller]

Nice!

 

Yeah, that gating system is really rough.  Smoothing that out will help.  

Edited November 23, 2022 by Jerrod Miller

[Jerrod Miller]

Another fun trick is adding some scratch vents.  On the parting line just scribe the smallest line from the part to the edge of the flask.  The goal is to make it easy for air to get out, but too small for metal to actually get in it.  I doubt they will be needed for your setup given the open vent via the shank.  

[Jeroen Zuiderwijk]

16 hours ago, Aiden CC said:

It makes sense that they would want to make parts like these hollow, especially since a casting needs some cleanup anyways. I was able to find a more zoomed in photo of those sheaths from above, and it does look to me like the boss and rim are different components on at least on the middle sheath, due to the shadows you can see in the flutes where they meet the central boss. Maybe a hollow, flat, head for the rivet and a thin sheet washer fluted with a punch and die or repousse? I think that the buttons on at least the bottom sheath could be formed from a single piece as they don't seem to have that same super sharp transition between the boss and the rim. Since I am still very new to casting, I think I'll probably start with thicker sections for the fittings I use to make it easier on myself while experimenting with other methods on the side. I'll try out casting the rivet and washer as separate pieces as well.

 

That does indeed look like two parts. I stand corrected  

[Jeroen Zuiderwijk]

5 hours ago, Jerrod Miller said:

Nice!

 

Yeah, that gating system is really rough.  Smoothing that out will help.  

I agree. The gating has to be as smooth and compact as possible. At every edge of the sand, the metal will pick up oil that will result in bubbles in the cast. You won't see that on the outside, but when you start filing or coldworking that's a problem. It's tricky enough as it is to get a casting with oil bonded sand that is solid enough. Rather then scraping channels, I use a smooth antler point to press the gating, and make very sure that the gating is completely smooth with no sharp edges or loose bits of sand. 

[Alan Longmire]

A few people have asked me about the articles I sent Aiden, so rather than continue using storage space for multiple PMs, here they are.  I think I got most of them from Google Books, but they are all available for educational use/creative commons copyright.

 

One of the most important bits of info is in the Technology of Medieval Jewelry, in which they state "casting as a technique of making finished objects is almost unheard of."  Ingots and rods, yes.  Other things, no.  Building up from sheet and wire is by far the most common technique, and soldering is far more important than it is now.

 

 

Anglo-Saxon jewelrywork.zip Fregni_The_Compleat_Metalsmith-libre.pdf The_Technology_of_Medieval_Jewelry.pdf The_wood_of_Merovingian_weaponry.pdf

[Pieter-Paul Derks]

I must admit that I have less knowledge of these things than some of the other posters in this thread, but I have done similar castings in delft clay.

Your second attempts look real nice already.

 

Personally I would cast this from the top, because it is the easiest surface to clean up again, and it would make it as easy as possible for the metal to fill the mold.

I made a quick ugly sketch to show what I mean, the red lines are air vents, I poke these with a thin bit of wire through the sand.

 

[Jerrod Miller]

For a simple button shape that is workable, but with the file work originally present that would leave anything that floats (sand or slag) stuck in those neat crevices.  I would also worry about pouring straight down onto the shank, as the sharp edge transition may erode, and that sand can go somewhere unpleasant.  Lastly, if your sprue isn't quite right you are more likely to get shrink in the center of the face.  Whenever possible important features (detailed faces) should be kept in the drag (bottom, pointing down) and away from gating.  

[Aiden CC]

Peiter, I may try that out, at least for plain rivets. Cleaning up the gate from the back is a little tricky. 
 

Jerrod, thank you for the tip about putting details in the drag. I’m looking into some non-casting routes for the washers, but I will still probably try out a cast washers and a one-piece casting too. 
 

Very quick and dirty, but the actual features came out pretty clean. With a simple jig I think I could make these pretty consistent. I’m still not sure about the best way to solder these pieces together, maybe a few bumps or shins in middle? Any advice there would be appreciated as well. 

[Jerrod Miller]

1 hour ago, Aiden CC said:

I’m still not sure about the best way to solder these pieces together, maybe a few bumps or shins in middle? Any advice there would be appreciated as well. 

Do they have to be soldered together?  The shank will go through the "washer", then when peened it will all hold together, no?  

[Aiden CC]

28 minutes ago, Jerrod Miller said:

Do they have to be soldered together?  The shank will go through the "washer", then when peened it will all hold together, no?  

That would work, and maybe be easier, though I have some lazy (and maybe some good) reasons to try soldering. 
 

The first is that a slightly off-center rivet in the washer is very evident. This means that either the hole and shank need to be exactly centered or one needs to be oversized. I’ve had a hard time getting the shanks square and centered since I poke through with a rod after taking out the model rivet. Would using a drill press for this maybe work better?
 

A second, (slightly) lazier, reason is that I found a hole saw is a super consistent way to make discs with a centered hole. The pilot is bigger than an acceptable rivet shank though. Also, I have been playing around with bending rings out of pattern wire, which would definitely need to be soldered. I’ll try building a jig for making the dimples and see if that helps with centering though, since it will give a very obvious circumference to grind and file down to. 
 

A slightly more valid reason is that since the washers will have a few points of contact, some “settling” could lead to a wiggle. If I sunk three high bumps in the middle or made three around the edge a bit taller, I could potentially get around that by really controlling the contact points though. 
 

I also want to try some decorative tinning like you see on some originals (could be a good way to make bronze rivets match brass washers too ), that might basically make for a free soldering step, maybe with something like tinning flux?