But is the process repeatable?

But is the process repeatable?

 

Based on the interest and comments that these articles have been receiving it seems prudent to offer up more in depth content. The focus here is the “how to”… the raw data will come later.

Sprueing:

 

Old School

Sometimes it easier to manually sprue. This depends on the easy of printability weighed against the complexity of the required sprue system.

Sometimes it easier to manually sprue.
This depends on the ease of printability weighed against the complexity of the required sprue system.

 

New School

Design your parts pre-sprued with easy to remove printable supports. This will save time, and minimize the labor post processing.

Design your parts pre-sprued with easy to remove printable supports. This will save time, and minimize the labor post processing. (STL link here)

 

How do you sprue? Well, to quote a friend and mentor ” You are a little silver car… that can only down shift… with no brakes…  and you are running out of gas.  Ready?  Pick the route to your destination during rush hour traffic.”  Now, imagine you are every car in traffic, that sums up how metal flow behaves. So, you build highways and roads to handle the traffic. Six lane highways with smooth merge lanes, soft lefts and rights for off ramps, no U-turns, wide intersections, and roomy shoulders for nervous drivers to pull over. You want to build clean effective roads for your frantic drivers on a hectic rush hour day.

 

 Theory meets Practice.

Theory meets Practice. (STL link here)

 

 

 Flask Choice:

Make sure the sprue tree doesn't exceed the height of the flask

Make sure the sprue tree doesn’t exceed the height of the flask.  Leave yourself 1/2 inch (13 mm) minimum clearance.

 

The flasks can be packed very tightly to get more parts with less waste, but be sure to leave enough clearance to prevent flashing.

The flasks can be packed very tightly to get more parts with less waste, but be sure to leave enough clearance to prevent flashing.

 

Investing:

These are the instructions for the investment being used "Satin Cast 20"

These are the instructions for the investment being used “Satin Cast 20”

Note: the size of flask can change the recommended burnout. Ranging from a 5-12 hour burnout depending on size and number of flasks.

Fill, the flask with 5/8 to 3/4 full with room temperature water. Then pour the water into a rubber bowl and invert the flask to drain.

Fill, the flask with 5/8 to 3/4 full with water [room temperature]. Then pour the water into a rubber bowl and invert the flask to drain.

1. Add investment until it forms an Island 2.  Mix thoroughly 3.  Check the consistency 4. De-gas/ De-bubble (this can be done by tapping, vibrating table, or vacuum chamber).

1. Add investment until it forms an Island
2. Mix thoroughly
3. Check the consistency
4. De-gas/ De-bubble
(this can be done by tapping, vibrating table, or vacuum chamber).

 

Pour the investment into the flask. Do not vacuum hollow printed parts,  rock the flask back and forth to eliminate larger bubbles, and top the flask off,.Then either tap the flask with the spatula or use a vibrating table to remove any remaining air.

Pour the investment into the flask. Do not vacuum hollow printed parts, rock the flask back and forth to eliminate larger bubbles, then top the flask off. Then either tap the flask with the spatula or use a vibrating table to remove any remaining air.

 

Burnout:

After the kiln is loaded the ramp up 200-250F should be slow until all the water evaporates from the investment. Normally a kiln will have a huge lag as it tries to get all that thermal mass up to temperature. The rate limiting steps is water's high specific heat, and poor heat transfer.

After the kiln is loaded the ramp up to 200-250F should be slow until all the water evaporates from the investment [see investment instructions (above) for burnout timing or burnout schedule here.] Normally a kiln will have a huge lag as it tries to get all that thermal mass up to temperature. The rate limiting steps is water’s high specific heat, and poor heat transfer.

 Between 450-600F the PLA is viscous enough to flow, but not burn. Scrape out the excess from the drip tray, and reload into the kiln. Let then kiln idle at this temperature until tray no longer has obvious signs of PLA. Scraping is done 2-5 times depending on how much  PLA was invested. If the parts were weighed before investing then the recovered PLA can be check by weight before continuing the burnout.

Between 450-600F the PLA is viscous enough to flow, but not burn. Scrape out the excess from the drip tray, and reload into the kiln. Let then kiln idle at this temperature until tray no longer has obvious signs of PLA. Scraping is done 2-5 times depending on how much PLA was invested. If the parts were weighed before investing then the recovered PLA can be check by weight before continuing the burnout.

 

Casting:

At this point photos get hard to take. Below there are links to video of the process (it is 30+minutes of video). It was kept in realtime so people will know what to expect.

First video is the casting of Wades Stack from the Rapid Manufacturing MK1 article. The metal ~1400F which is excessively hot, but it seemed prudent to show a workable temperature range.

1st video Wades Stack     (Video Length 10:00 minutes)

Feel free to skip ahead to the time stamp in the image

Feel free to skip ahead to the time stamp in the image

 

Second video is the casting of the  Gyro Ornament in the sprueing example above. The metal is cast at the ideal temperature, this is more important for thin feature parts, and heavy parts, both are more drastically affected by linear contraction from high temperature cooling.

2nd video Gyro Ornament   (Video Length 9:12 minutes)

Feel free to skip ahead to the time stamp in the image

Feel free to skip ahead to the time stamp in the image

 

Third video is the quenching process. Remember, aluminum and plaster hold a lot of thermal mass, and erupt violently in the water. When it is done correctly it is safe and quite anticlimactic.

3rd video Quenching   (Video Length 6:47 minutes)

Feel free to skip ahead to the time stamp in the image

Feel free to skip ahead to the time stamp in the image

 

Fourth video is temperature data for the proper tempering and quenching for the aluminum. Any surface defects are pointed out in the castings. And, there is a quick demonstration on de-sprueing the wades stack.

4th video Tempering/De-sprueing    (Video Length 8:02 minutes)

Feel free to skip ahead to the time stamp in the image

Feel free to skip ahead to the time stamp in the image

 

 

This is was cast in the video. It was the third iteration of sprueing. The sprue tree photographed was effective.

This is was cast in video 4. It was the third iteration of sprueing. The sprue tree photographed was effective.

 

Fifth video is of the working Gyro Ornament… After de-vesting, de-sprueing, some clean up, and tweaking.

5th video functioning Gyro Ornament (Video Length 20 seconds)

Feel free to skip ahead

Feel free to skip ahead

 

 Wade's gears after the stack has been de-sprued.(Wades Stackable link here)

Wade’s gears after the stack has been de-sprued. (Wades Stackable link here)

 

 

Troubleshooting and Diagnostics: 

 

Symptom: Partial plaster embedded in the part.

Possible Causes: Plaster shift, Porous print, Flashing, Checking, Hollow voids

Hexagonal patterns (Left) are signs that the printed part was not water tight. Large flaking (Center) are signs of poor flow between thin and bulk areas of a part, commonly followed with more spalling. Slitting and Spalling on (Right), is a sign that the contour could not withstand the pressure of the metal flow. Either the plaster was mixed to thin, was physically too thin , or a combination of both.

Hexagonal patterns (Left) are signs that the printed part was not water tight. This is the disadvantage to vacuuming the part in plaster.  Large flaking (Center) are signs of poor flow between thin and bulk areas of a part, commonly followed with more spalling. This can be better solved with proper sprueing. Slitting and Spalling on (Right), is a sign that the contour could not withstand the pressure of the metal flow. Either the plaster was mixed too thin, the part was physically too thin, or a combination of both.

Diagnostics:

Hole diameters in the part are either too small, or too long to withstand the force of the metal flow [TestFeaturesCast.stl].

Hexagon formation, small hexagonal recessed portions indicate that the surface of the 3D print was not water tight, and plaster leaked into the interior geometry.

Air bubbles commonly form on interior surfaces or convex surfaces relative to the parts orientation, allowing fragments of plaster to shift easily at the meniscus’ edges.

Solutions: Check investment mixing instructions, make sure printed part is water tight, do not vacuum part, redesign sprue systems, minimize thin features under  1/16″ or [1.4mm].

 

Symptom: Cracking

Possible Causes: Premature quenching, large volumetric contraction, or both.

Volumetric contraction tends to cause cracking and shearing internally when right angles from at points where the bulk metal have extreme differences in the surface area to volume ratio.

Volumetric contraction tends to cause cracking and shearing internally when right angles from at points where the bulk metal have extreme differences in the surface area to volume ratio. (VolumetricShrinkTest STL link)

 

Diagnostics: If the part is heavy it is most likely due to volumetric contraction, this is most common in aluminum.

Solutions: Fillets and chamfers help diffuse these internal stresses. Redesign with gussets, ribs, piercings, through holes, or webs instead of solid mass parts.

 

Symptom: Bulging/Blebbing

Cause: Premature quenching

These are two different views of the same part.  Bulging is the precursor to Blebbing, they are cause by steam expanding into the molten metal.

These are two different views of the same part.
Bulging is the precursor to Blebbing, caused by steam expanding into the molten metal.

Diagnostics: Bulging is the precursor to blebbing. It occurs when the molten metal comes into contact with water. The energy transfer of water turning to steam forces the steam to expand into the molten metal pushing it out through the hottest remaining contours.

Solutions: Wait longer before quenching, use a thermocouple if necessary to measure the temperature. Aluminum remains viscous ~900F. You can temper the aluminum by quenching between 500-750F. Remember the heavier areas of a part will take significantly longer to cool, the thermocouple will only give you a vague average of the button temperature. Bronze similarly will stay viscous ~1500-1600F …same rules apply.

 

Symptom: Splicing

Possible Causes: Failure to print a continuous layer, bad flow in mold, or crack in the plaster, firing the kiln too quickly can cause the PLA to expanded and crack the plaster in one continuous plane.

Splicing is often a plane shearing in the printed part, the plaster or both.

Splicing is often a plane shearing in the printed part, the plaster or both.

Diagnostics: Reprint part and look for layers of delamination, make sure the extruder can run continuously. For flow issues, test the edge cases with TestFeaturesCast.stl  the pitch example can ensure that the metal is flowing properly through the mold.

Solution: Minimize the thin features under 1/16″ or [1.4mm].

 

 

Tags: , , , , , , , , , , , , , , , , , , ,

9 Comments on Rapid manufacturing MK3: Detailed Walkthrough & Troubleshooting Guide

  1. JULES says:

    Keep it up and we’re going to win!!! Here is a gear I made. http://i1269.photobucket.com/a.....d9354f.jpg

  2. […] For the past year we have been busy building, testing, documenting and refining the process of taking 3D printed parts and using “Lost PLA” burnout to cast for parts for more robust applications. The documentation is bordering 100+pages, with 20+ pages of brute force data. We will try to keep it simple, show off with a few shiny throwbacks, hopefully inspire ideas for the potential, and give some technical specs to boost the capabilities of those open source open hardware folks who love a good clean walkthrough. [click here for more information] […]

  3. aonemarine says:

    Bowman, contact me at aol.com

  4. Bill Hurley says:

    I have been exploring the possibility of direct casting of silver and brass into graphite molds for small scale production runs. The molds I would like to create have surfaces that are to complex to be machined on a CNC, which leads to 3D printing. What if there was a way to utilize 3D printing for creation of graphite molds for direct casting of metals? A negative mold, like those used in sand casting, could be 3D printed in PLA or even wax, the mold could then be filled with a graphite slurry consisting of graphite powder and adhesive. The mold could then be either separated physically or burned off and you would have your graphite mold. This would have to be done for each half of the master graphite mold. The surface of PLA mold could be coated with one of the graphene inks/paints that have been recently been developed (reference links below) to address the issue of the graphite slurry sticking to the PLA mold. No equipment would be clogged up using this method and it would be fairly inexpensive. What do think, am I crazy?
    http://www.aremco.com/news-item/1486-2/
    https://graphene-supermarket.com/Conductive-Graphene-Dispersion-100.html
    http://www.telegraph.co.uk/fin.....stuff.html

    • bowman says:

      Printing in graphite is possible. This wouldn’t be a good FDM/FFF 3D printer method. Graphite printing would be much better on a powder printer, where and the inkjet head runs the binder/adhesive across the bed of graphite powder. We had good luck in the lab pouring bronze directly into powder printed hydroperm molds.
      http://open3dp.me.washington.e.....hydroperm/
      Simply swap out the hydroperm with graphite. Make sure the wall-thickness/binder ratio can handle the metal-load and the thermal-shock. If the binder is robust it should even be able handle a light vacuum load.

  5. Bill Hurley says:

    Thanks for that, but I am still curious if you have any thoughts on the use of graphene paint as a coating for the molds.

    • bowman says:

      My apologies, I failed to clarify that earlier. I was suggesting that graphene is an extra step that could be automated in a powder printer…
      To answer your question…
      If you are running and FFF/FDM setup the graphene paints should work fine. Graphene is carbon so it will go straight to CO2. I have used pure graphite as a release agent for high temp forging/pressing and die striking applications, and it works well. The carbon shielding [graphene] will act as a release agent,a major thermal insulator, and a high reduction atmosphere… meaning the castings may take longer to cool down. If you are casting brass that is generally ~30% zinc which will oxide readily in a carbon rich atmosphere [Aluminum can have similar problems]. Stick to using silver, bronze, shibuishi alloys.

  6. Dennis Meyer says:

    Do you have recommendations on the print density, shell layer count or slicer?

    • bowman says:

      All of the example castings were printed with these standards:

      Makerware
      .2mm layer height
      2 shells
      8% infill

      This standard offers enough strength-to-weight ratio and maintains a print that is water-tight. Although any number of slicers can be used and are effective, MakerWare was the default for the Replicator 2 printer used for testing.

      Slicers that work well are:
      Cura
      Simplify3D
      MakerBot Desktop
      Slic3r
      Replicator G (takes forever but generates a good toolpath)

Leave a Reply