Metal 3D print

[GreenTwin]

It is such a paradigm shift from 2d

This is true, but I also learned 3D a few years ago, after 32 years of 2D work, and 8 years of hand drawing work prior to that.

For 3D work, one has to think one part at a time (for the approach I use).
Its like slicing an apple; you see a circle on the perimeter of the cut.
So work backwards; draw 1/2 circle, rotate it around an axis, and presto, you have an apple.

For every part, I think/plan on where I could take the most effective cut to begin the sketch, and then add onto or subtract from the solid shape I create.

At first, I was thinking in terms of "drawing an engine".
Once I figured out that I should model one part at a time, and then assemble the parts later, I figured it out.
.

 

[Gordon]

I bought Alibre Atom last year and I watched a bunch of videos and started making drawings of an engine that I had 2D drawings for and spent about a week working with it. I had also used 2D cad for about 30 years and that was a problem because the 3D is a whole different concept than the 2D. I am now at the point that I can make a 10 minute sketch in about an hour. Lots of new tricks to learn. I am 84 so we are never too old to learn something new and at 84 I no longer have to worry about making money with it.

 

[Lloyd-ss]

Come on Lloyd, I learnt Fusion to a useful level a couple of years ago, at the age of 77. It's nowhere near too late. I suspect there are plenty of teaching videos on Youtube. Going back to Autosketch is like stepping back into the dark ages!

This is true, but I also learned 3D a few years ago, after 32 years of 2D work, and 8 years of hand drawing work prior to that.

For 3D work, one has to think one part at a time (for the approach I use).
Its like slicing an apple; you see a circle on the perimeter of the cut.
So work backwards; draw 1/2 circle, rotate it around an axis, and presto, you have an apple.

For every part, I think where I could take the most effective cut to begin the sketch, and then add onto or subtract from that solid shape.

At first, I was thinking in terms of "drawing an engine".
Once I figured out that I should model one part at a time, and then assemble the parts latter, I figured it out.
.

Well!
It looks like I got a couple of boot prints in the butt...... and some age shaming for saying I was too old to force myself to learn Alibre 3d.

And @Gordon , I see you just posted to make me sink even deeper into the chair. No excuses allowed, I guess. I agree that nothing is faster than a pencil sketch, but the cadd work sure saves on mistakes at the machine.
Lloyd

 

[GreenTwin]

I learned by thinking in terms of the old Playdoh extruder.
You are basically drawing the sketch of the extruder die, to either extrude a solid, or extrude a cut.

The other thing that stumped me for a long time was what I call "plane jumping".

In 2D, you just draw lines on a flat sheet of real or virtual paper.
You always remain on that sheet, on that single plane.

With 3D modeling, you get onto either a front, side, or top plane (or a custom created plane), make a sketch, and then extrude a shape.
One you have extruded a solid shape, you default back to not being on a plane.
To make the next step, you have to decide which plane you want to get on, select that plane, make your sketch, and extrude or cut on that plane.

So you are jumping on and off of various planes; that was very confusing for me coming from the 2D world.

Then I had to figure out how to create new planes in the positions I wanted them, and this is done by offsetting from one of the basic three planes (front/top/side) some determined distance.
Often the newly created plane is offset and at the same time that it is rotated.

It took me several months to wrap my head around finding an efficient starting shape/sketch, and then placing additional planes around to add/subtract from the initial shape.

Once you get the concepts down, 3D is actually pretty easy.

.

 

[Gordon]

The biggest problem is trying to do too much. With 2D you would draw all of the holes on the sheet regardless of size but with 3D you take it in smaller increments. Like Green Twin said you start out with a basic shape and extrude added or subtracted. Like making a statue. Start out with a block of marble and take away everything that does not look like a horse.

 

[f2cf1g]

Returning to 3D printing, here is better evidence of the print orientation. This is a macro picture of the top of the block in oblique light. It shows that this surface has shrunk unevenly across the whole face. In contrast, the vertical surfaces at either end of the crankshaft housing show no signs of shrinkage. If the cause is gravity the print was made cylinder up, not from the backplate up.

Logic suggests that the laser energy is being dissipated into the surrounding metal powder faster at the surface, which contracts less than the core where the slower rate of energy dissipation allows the particles more time to fuse and contract. Presumably, therefore, the density and thus other material properties vary across the section. It will be interesting to see whether this has any practical impact. There is a machining allowance on this surface so the unevenness is no problem.

 

[Richard Hed]

Well!
It looks like I got a couple of boot prints in the butt...... and some age shaming for saying I was too old to force myself to learn Alibre 3d.

And @Gordon , I see you just posted to make me sink even deeper into the chair. No excuses allowed, I guess. I agree that nothing is faster than a pencil sketch, but the cadd work sure saves on mistakes at the machine.
Lloyd

Ha haa haawww. You guys are too funni. A 2D sketch is faster than CAD? Well maybe, if you are using some 2D sketchingh system from the last millenium but if you are doing 3D, the drawing is actually MUCH faster. 3D is faster and EASIER than 2D, but it may take a bit of learning the system.

Gordon: what's the problerm? I ahve to admit, Atom 3D at first confused me because of the need to "clik" on the co-ordinate plane to start. After I got that, it was just downhill liearning. Are you still having trouble?

 

[Gordon]

Gordon: what's the problerm? I ahve to admit, Atom 3D at first confused me because of the need to "clik" on the co-ordinate plane to start. After I got that, it was just downhill liearning. Are you still having trouble?

I have learned a lot. I still keep learning new tricks which make life easier. One of my big problems is when I place a new shape and I want it to reference to the origin I end up placing it too close to the origin and it may look like it is at the origin but it is a few thousands off so I end up trying to dimension it to "0,0" and I cannot get the dimension to show. I am sure that I am doing something wrong since others so not seem to have that problem. Actually the dimensioning on Alibre is the thing that I find the most frustrating. When trying to fully define a part I end up with dimensions all over the place trying to figure out which dimension is missing.

Enough of hijacking this thread.

 

[Bazzer]

Returning to 3D printing, here is better evidence of the print orientation. This is a macro picture of the top of the block in oblique light. It shows that this surface has shrunk unevenly across the whole face. In contrast, the vertical surfaces at either end of the crankshaft housing show no signs of shrinkage. If the cause is gravity the print was made cylinder up, not from the backplate up.

Logic suggests that the laser energy is being dissipated into the surrounding metal powder faster at the surface, which contracts less than the core where the slower rate of energy dissipation allows the particles more time to fuse and contract. Presumably, therefore, the density and thus other material properties vary across the section. It will be interesting to see whether this has any practical impact. There is a machining allowance on this surface so the unevenness is no problem.

View attachment 157531

We did not get any such problems on the NB 40 cases done by DMLS process. We are assured that the material is homogenous and consistent throughout.

Accuracy is very good as well.

 

[Richard Hed]

I have learned a lot. I still keep learning new tricks which make life easier. One of my big problems is when I place a new shape and I want it to reference to the origin I end up placing it too close to the origin and it may look like it is at the origin but it is a few thousands off so I end up trying to dimension it to "0,0" and I cannot get the dimension to show. I am sure that I am doing something wrong since others so not seem to have that problem. Actually the dimensioning on Alibre is the thing that I find the most frustrating. When trying to fully define a part I end up with dimensions all over the place trying to figure out which dimension is missing.

Enough of hijacking this thread.

Oh. I have had the same problem--it often turns out that I have a dimension that needs to be removed which stops the dimension you need from being realized. Of course that is called "over dimensioning". The machine will not allolw that

 

[Lloyd-ss]

Returning to 3D printing, here is better evidence of the print orientation. This is a macro picture of the top of the block in oblique light. It shows that this surface has shrunk unevenly across the whole face. In contrast, the vertical surfaces at either end of the crankshaft housing show no signs of shrinkage. If the cause is gravity the print was made cylinder up, not from the backplate up.

Logic suggests that the laser energy is being dissipated into the surrounding metal powder faster at the surface, which contracts less than the core where the slower rate of energy dissipation allows the particles more time to fuse and contract. Presumably, therefore, the density and thus other material properties vary across the section. It will be interesting to see whether this has any practical impact. There is a machining allowance on this surface so the unevenness is no problem.

View attachment 157531

Do different vendors use different materials, machines, methods, etc? I guess they do(?)
But still all of the castings printed parts pictured in the thread look extremely nice, as good as investment castings but at a much, much, lower cost. I will be interested to see if all the printed parts are free of internal defects. If they are good dimensionally, internal defects can show up where there are extreme thickness variations in a castings, but maybe not in printed parts??
Overall, extremely nice, and honestly, I am a bit amazed and pleasantly surprised.
Lloyd

 

[f2cf1g]

Picking up this thread again after being away, the comparison between SLM and DMLS is interesting. According to the PCBWay web site, the two processes seem to differ only in the degree of melting of the metal particles. DMLS melts them partially, leaving some rigid material to provide support but with the danger of porosity, whereas SLM melts the lot, making it more solid but dimensionally less precise. That would tally with what appears to have happened at the top of my print. Bazzer's DMLS print doesn't seem to suffer from porosity so maybe ultimate quality is down to the individual company and its specific implementation of the process. PCBWay seems not to offer DMLS but I found a UK company which does, unfortunately at six times the price.

 

[Jasonb]

I suppose it is a bit like the old traditional methods, yours is what might be expected from a good sand casting whereas Barrie's crankcase would have needed an investment casting or die casting to get the detail which would tend to put the cost up particularly if it were a short run.

Again like sand casting the first pour may show places where there has been shrinkage so a pattern may need modifying to have a larger machining allowance or the gating altered.

 

[Bazzer]

We will be running a job at a commercial print house, there is some space on the build plate maybe you could jump in on that space? You can see the space at the top of the build plate.

 

[Basil]

Very interesting. What is the timeline on this and what is the rough cost of the prints you have on the table each. Thanks

 

[Bazzer]

Very interesting. What is the timeline on this and what is the rough cost of the prints you have on the table each. Thanks

What is on the table (6 parts) is £1200

Timeline is probably 6 to 8 weeks for submission of final design.

 

[Basil]

Thanks Bazzer. I'll contact you directly on this.

 

[petertha]

@Bazzer is that angled print orientation in post#34 picture something that the 3DP software sets up for its own optimization?
What engine displacement sizes are we looking at just for layout scaling?

 

[Bazzer]

@Bazzer is that angled print orientation in post#34 picture something that the 3DP software sets up for its own optimization?
What engine displacement sizes are we looking at just for layout scaling?

No the angle orientation is setup by the company who is doing the work, this angle is according to many factors mostly relating to support of over hangs.

The big cases (4 off) are 6.5 cc.

 

[petertha]

We will be running a job at a commercial print house

Did you select this particular print house (over other online shops) more for what you deemed to be their technical expertise, end product quality, prior experience etc.? Or maybe it was more driven by home proximity / timing / cost...? Is there a batch cost advantage to filling up a specific volume, or that engine size/count just so happened to meet your requirements for now? Do you think there any particular concern for 3DP aluminum heads? (specifically thinking about elevated temps & cycling temps). This is exciting stuff!