Moving parts WSF clearances ?

I have a question about the WSF and clearance.

From the folowing page :Creating-hinges... I got this info :
However, when reading the forum, experiments by users show that at least a clearance of 0.5 to even 0.75mm is apparantly needed for WSF ? Were these just flukes or what ?

I have some designs in queue for production, with moving parts and clearance around 0.3 mm.
Gyro the Cube and Twirly pendant
Will they print ok ? Or do they need redesign or special care ?
Since they are not ordered by me (yet) but by a customer, I'd hate to disappoint them with flawed designs.

I wouldn't trust the SWF material for gaps under .6 mm. In the cases where I have experimented, .5mm gaps *fuse*:

-a ball and socket test

-my nocturnal

Both of these pieces use a .5mm gap in the tightest areas. In both cases they fused.

I'm ordering jdavidbush's piston test ... actually it will probably arrive tomorrow. I ordered it in transparent detail and in SWF materials. I'm hoping to see just how big of a gap I need in SWF, and just how small of a gap I need in the detail materials. I may order these a second time, to be sure that printing flukes don't easily alter the gap needs. Then I would have an experiment with 3 reps (including jdavidbushe's results) to rely on.

-Whystler

Ok, thanks for the reply.
I read alot in the forum about these clearance problem, but thought those were all due to batch errors..

I also understand that the clearance has to be bigger if the surface is larger, which is understandable; powder could get locked inside, like sand in a ball-bearing.

But using the air gap design from the christmas ornament and the clearances from the creditcard gear test, I thought was way in the clear.

I've used 0.5 mm clearance for my scale:



But maybe I was just lucky, or it was a combination of the shape and printing direction. At first it didn't move and I had to move it strongly to make the powder go out of the gaps, but then it was fine.

Hey Frank,

What you have there is very interesting.

I see a lot of ring shapes that remind me of a torus there. If, at the smallest point the opening in the middle of the torus is .5 clear from the inserted axel, then I can see why your design worked with SWF.

Its because the .5 clear only happens on the edge of the curve for a minutely small distance. If, instead, you were using a tube, where the distance was uniformly .5 clear for even a mm or two, my experiences would make me think that it would have been fused.

This is actually good news for many folks. If you are building an axel, and you want to get really close using SWF, then make sure the rings that hold your axel are tori, and not tubes. I wonder if you can actually get closer using this technique?

-Whystler

The distance was 0.5 mm for the whole length of the cylinder, but an image of a hinge cross section can explain it better than me



As far as I can see, printing direction was as like the scale was standing upright on its three feet in the printer.

Well, then it was strangely unfair that it worked for you! lol

It makes me wonder if I should try and print the nocturnal again in SWF...

-Whystler

OK. Let's try another hypothesis:
- Perhaps the round part of the design helped the flow of air to enter into the small space between the different parts, during the cleaning process. In other words, it may have concentrate the air flow exactly where it needed to go.
What do you think?
Should we chamfer all the edges?

Hi guys,

I've got a couple (somewhat long winded) examples that might help illustrate some of the difficulties in determining the what the proper clearance should be:

Example 1:

A 2mm diameter axle passing through a 3.2mm diameter hole (leaving a 0.6mm gap on each side). The axle was printed horizontally ("laying down").
If I measure the axle at (A), it measures between 2.09 and 2.20mm (it is larger in the "vertical" print direction.)
If I push the axle through the hole and measure it at (B), it measures between 2.23 and 2.30mm. About 0.1mm larger.

Example 2:

A 5mm diameter axle passing through a 6.2mm diameter hole (again, a 0.6mm gap on each side). The axle was printed vertically ("standing up").
If I measure the axle at (A), it measures between 5.01 and 5.03mm. If I push the axle through the hole and measure it at (B), it measures between 5.13 and 5.15 mm. Also, about 0.1mm larger.

The measured dimensions in Example 2 are more accurate than those in Example 1. (And the "range" is much tighter.) I believe this is mostly due to print direction.

You can see why a gap of 0.3mm would be too small in Example 1. If the axle is 0.3mm too large, the hole it passes through is most likely 0.3mm too small. (I have to *assume* that it is, as I'm unable to measure it. However, I'd be perfectly happy cutting the model to make that measurement if anyone thinks it would help.)

Also, in both examples, the section of the axle that was printed inside the hole is larger than the sections outside the hole. Here is an example of what I *think* is happening... when the laser sinters the inner cylinder (the axle), excess heat dissipates into the "gap material" and it heats up; just not quite enough to melt. Then, when the laser sinters the outer cylinder, the gap material heats up a little more and the some of the material (especially the power right next to the cylinder surfaces) melts and fuses.

Okay, I've talked enough. Let me just say, I have to agree with Whystler. I've also printed a few parts with 0.5mm gaps that were fused. (In fact, the only gap I've been able to print that was smaller than 0.6mm and not fused was a 0.43mm gap between the teeth of two gears. So, congratulations, Frank! )

~David

Hi David,

This is interesting indeed. But if the axle is larger when it is at proximity of the hole, can we also suppose that the hole gets smaller : so, according to you, would we have to add 0.1mm to te radius of the axle and substract 0.1mm to the radius of the hole a margins, in order to completly take into account this phenomenum?

Just to chip in; I've had a design featuring extremely fine text printed all non-metal options bar cream robust. The results were interesting.

In order of greatest 'quality' or accuracy:

1. Transparent Detail. Good reproduction.

2. Black / White Detail. Poor. When sanded and cleaned with fine needle, some text is visible.

3. WSF. Unusable. Edging of text is 'clumpy', where loose powder has obviously fused during cooling, and standing an appreciable distance from the otherwise planar surface. Even major internal modelled voids varied in quality; one was rapidly cleaned with an old electric toothbrush (really speeds this part up) followed by a needle, the other would have to have been carved to the desired shape.

To conclude, WSF tolerances are rather variable, especially where there exists the possiblilty of additional unwanted powder to fuse in place. I'd hoped to make a master pattern in this material and perform some carving and cleanup, but will probably be staying with Transparent Detail.

To be honest, I knew in advance I'd be pushing my luck with the WSF, but was surprised by the poor definition, even given the miniscule detail I was trying to have printed.

I have some interesting news ...

Jdavid posted above using the example of his test piece, which is available for sale. In fact the test piece is a row of those axels using different clearances.

I ordered it in transparent detail, and SWF, and it arrived today just in time for this discussion Playing with it heavily (twisting and pushing) to dislodge support material in both models, I soon found that his obervations were true to my pieces as well ... those being:

SWF would move with .6 clearance.

detail would move with .2 clearance.

BUT THEN ...

I tried *REALLY* hard with the .5 clearance on SWF, and *POP*, the axel pushed through, and is now clear. It, like the others, was jammed by support material. It was just a lot harder to budge it.

I continued my brutish twisting with .4 clearance, and bent that axel a little. So, I will say that .4 clearance for those axels won't work.

I would still shy away from .5 in most situations for SWF. Because I did have to push, twist, and pull very hard to get the axel out. If you have a delicate or detailed model, you might cause it damage.

-Whystler

Ok guys, wow thanks for all the informaiton and sharing of experience !!

I must say, I'm kind of disappointed, I will need to do a recheck and redesign on many of my models

Did WSF become less accurate over time, or what happened?
The above mentioned gear card worked properly with 0.25 mm back then, apparantly without special measures. So why doesn't it now ?

Most of the the clearance and axle design I see are using large closed "contact" surface. How about the hinge design from the christmas thingy, it uses 0.5 mm with extra air gaps (reducing contact surface), As I understand from Whystler's experience, these could probably still work, as long as there is enough room to wiggle out any loose powder ? Do people have experience with "open" hinges like this ?

I will be getting to it soon myself, hope to design and print lots of test parts, to figure it all out.

Apparently the SW&F resolution variability is related to powder recycling, so to provide better predictability, I would like two levels of SW&F quality. One inexpensive type made from mostly recycled material for low res prints, and a slightly more expensive SW&F from all virgin material for detailed prints.

This would provide the best value for people not to concerned with detail, and ensure accuracy for those who are.

Also, I have noticed, as Psau3 pointed out, that the sharpest results are coming from Transparent Detail. I think this is because of added pigments, the same way that paint loses quality when too much tint is added.

Material purity is now a recurring theme. Beyond print resolution, quality control will be the new challenge for good prints.