I think the fogger idea is pretty good! Id give that another go!

At some stage I'm going to experiment with using small fans to blow around the fumes. I've already tested one passively and the plastic wasn't affected by the acetone.

Here's what you need:

1) A large mason jar with lid
2) acetone
3) nothing else (except maybe some wire to hold the model)

While the vacuum set up did a good job of evenly depositing a thin layer of acetone onto the model, it only accelerated the smoothing process that could occur simply by leaving the model in the jar a little longer.

As before, I successfully tested smoothing a vacuum-treated object (froggy) by applying a vacuum to boil acetone and leaving it in the jar after the vacuum was released.  After an hour or so, the frog was appreciably smoother and evenly shiny all over.  As a control, I tried just putting the frog into the jar (with some acetone at the bottom) suspended by a wire and sealing the lid to see what would happen.  It took about two hours to get the level of smoothing the vacuum-treated frog got in one, but the smoothing continued to happen.  After leaving it overnight, it had become one totally amorphous blob (100% smooth!). 

As an extra safety measure, I kept a latex balloon attached to a bulkhead fitting mounted through the lid, and confirmed that there was no measurable pressure build-up at any time from the smoothing process.

So it seems that I've come back to the old standby paint-can method, except that using a sealed mason jar keeps the fumes inside and preserves the acetone so only very little gets used. 

Without an escape path for the acetone to evaporate away, I'm curious whether sealing the jar helps concentrate the acetone or make it more evenly distributed in the jar.  So my next test will be to see if I need a fan to circulate the fumes to prevent a gradient in vapor concentration from making the smoothing uneven.  I'll switch to the larger mason jar I had previously used as the top-chamber, and try smoothing a tall object instead.

I've got my rice cooker going and the results are pretty good, but I still want to design a better way, for a challenge as much as anything.

I have performed four different types of smoothing: A simple 5 second dunk in acetone, the vacuum chamber, acetone spray and heating the acetone (Presto Method).

Previously I mentioned about using an ultrasonic fogger to generate some smoothing but this turned out to be a bust straight away.
After getting everything set up I put the fogger in the acetone and nothing happened .
This was due to the level sensor built in to the fogger to make sure you didn’t run it dry. It used conductivity to sense the level and acetone isn't the same as water if you didn't know. However in the very short time it took to realise this was never going to work without building my own fogger the casing to the it had began to very quickly melt. I had to whip it out of the acetone and dunk it in water to try and stop the process before it fell apart, the housing had turned to rubber in about ten seconds. I might use this idea to get the electronics out of the housing yet, but at the time it was much more of a panic to try and save it. So anyway that idea was a none starter and I’ve left it alone for now.

For completeness as a comparison this is close up pictures of an unharmed frog

Starting with the simple 5 second dip:
As I had a jar of acetone and the frog fit quite nicely in the top a bit of copper wire around the poor thing was plenty good enough for a nice bath

That worked better than I expected, well once I realised that he was sticking to the surface I put him down on I manged to pull him up and let him air dry.

Next up was the vacuum chamber:
This is the setup I used with my hand pump.
And a quick one of him taking one for the team.

After a while I took him out, let him dry off and got these results

From the underneath picture you can see where the acetone did most of its work where the vapour has rushed in from the pipe which was directly below the model. The belly up to the throat is lovely and smooth. You can see some more subtle smoothing on his head and back. The whole model does have a gloss quality but the low resolution print was too much to over come in this case.
With a better print and a shield for the incoming vapour this could have had better overall results.
I have a problem going forward now as I've broken my hand vacuum pump, so need to get a proper electric version if I am going any further.

I then moved on to using an acetone spray via an airbrush. I had one that came with an old compressor but figured it was going to be designed for use with solvents so it should hold up well and most of the parts that touch the acetone are made of brass.
There was general problem with this though. As I had to use it outside the fine spray of acetone was just able to make the model wet before it dried so even though I tried for a while it didn't give very good results at all, leaving the model with a matt finish, although this could have been some contaminant in the airbrush as I had never used it before.

Finishing off I used a rice cooker that seems to be as close to the Presto cooker I could find in the UK at a cheap price.
I ended up with this...

I think you can see that the model is much smoother than previous methods. You can just make out some marks on the belly where the frog started melting so quickly that I had to loop some wire around him and pull him off the tray it was sitting on but by then he was stuck so I ended up getting the wire stuck to him.
This was my fault for underestimating how quickly this would work, next time I will try for a much shorter period before turning it off and getting him out to dry.

Some observations:

If you want to smooth more make sure air doesn't get to the model. Once the air is let in everything start to solidify quickly and although it doesn't harden for quite some time the smoothing process stops very quickly.

Using any method can get decent results if you keep trying and work around any issues you find.

Acetone isn't as dangerous as everyone seems to try and make out. Whilst airbrushing the frog I thought I would see how flammable the spray was and got a lighter out. The spray would just put the lighter out every time. So thinking the flame wasn't hot/big enough I got a blow torch out and found that just air would interfere with the flame as much when the acetone was coming out.
Now I won't say it is safe, clearly it is flammable, but when using the Presto method I just had it on a window sill with the window open. I can't see any possibility of any harm coming from that. If you are a smoker probably best not to smoke around it though.

Sorry for the long post, I tried to trim the pictures to the bare minimum, I had 160MB to start with.
For now I will be using the Presto method as it is easy, but I need to get a timer to get better control of it.

Aaaaaaaaaargh....I'm melting, what a world......

Testing complete, results to follow with flipin' huge pictures

I quickly printed these at .3mm, 2/3 scale.  Smoothing tests to follow.

Did you do those in slic3r, if so what version?  Can you complete the top of the head without any infill?

Boo, looks like we cant embed vine.

https://vine.co/v/bTJWjF9qj9Y

All printed at 0.3mm so the smoothing effect can be seen the most.

I wonder how long it takes to dissolve a frog?

Good point about the acetone landing on the model doing the work but I think as the fog condenses out out the air around it I hope it should be enough to make a difference, plus it looks really cool.

That is a much better result than I achieved.
I am now thinking that with my setup letting the air back in was reducing the amount of acetone in contact with the model as it was being absorbed in to the air and not being deposited on the model.

Since my last attempt I have now got an ultrasonic fogger which I plan to see if that generates enough acetone moisture to affect the model. Also need to find out if it will just melt in the acetone the electronics are encapsulated in resin but the case might still fail.

Awesome.  I can't wait to read the results of your new experiments!

As far as the fog goes, I've personally come to the conclusion that acetone vapor doesn't by itself seem to result in any smoothing, but that it has to condense to a liquid on the surface of the model first.  In a normal heated smoothing setup, the acetone is heated up while the model is at room temperature, so this causes the acetone vapor to condense onto the model as liquid.  In my vacuum setup, however, everything is at the same temperature, which is why I released the vacuum after allowing the acetone vapor/liquid to reach equilibrium.  This caused the vapor to need to transition back to liquid, so it condenses on every surface it can find, including the model.

I'm not sure how or if this might apply to a fogger, however, so I'm definitely interested in how this will go.  I'm very happy with my vacuum setup for small objects, but a fogger setup could be scaled up more easily, I think.

Since my last attempt I have now got an ultrasonic fogger which I plan to see if that generates enough acetone moisture to affect the model. Also need to find out if it will just melt in the acetone the electronics are encapsulated in resin but the case might still fail.

I'm currently printing some treefrogs at 0.3mm as test subjects so the pictures look a bit more interesting.

Now the in-laws have left testing resumes tomorrow, I now need more acetone.
Plenty of back to back testing to come, as well as seeing what happens when you light acetone fog

On the weekend I took one for the team and chucked in a spare 30mm fan to test if it would be affected by Acetone. THe plastic parts of the fan are fine, so I suppose the only thing that could have failed would be some of the bonding parts in the circuit. So far so good.

The design pretty closely follows the sketch I included in an earlier post.  Two mason jars (1/2 gallon and 1 pint) serve as vacuum chambers, with the part to be smoothed held in the smaller chamber, which is connected to the larger jar above.  A balloon in the upper chamber separates the air between the two chambers and keeps the acetone from escaping when a vacuum is pulled in the upper chamber.  I pulled a vacuum using a compressor-driven venturi pump that I got at Harbor Freight for $15.

As you can see from the attached video, the balloon expands when the vacuum is pulled, causing some of the acetone to boil off into vapor.  After only a few seconds, I released the vacuum, and acetone condenses evenly onto the part and the sides of the lower mason jar (which you can see as the fog that suddenly appears).  Since the lower container is sealed and the air in there is now fully saturated with acetone, the acetone doesn't evaporate and continues to work.  I then left the part to smooth for an hour, and the results are below.  The part had a nice gloss to it.  It was still sticky with acetone when I pulled it out, so the smoothing would have continued if I had left it in for a longer period of time.

I used an old discarded print as a test which had a fair amount of wobble to it.  About half the wobble was gone, so in the future I'd still want to use wobble-compensation and maybe some presanding, or leave it longer in the chamber.