Tell me about water ripples, waves and surface tension. Is this really an issue?

[InstantWork]

Hi fellas.

While reading some 2threads here I´ve come across the terms, surface tension, ripples and waves. Every time these are mentioned as a problem and different solutions have been discussed.

Can anyone inform me on how these really influence the prints, or if they do at all, can you please describe why they are still an issue?

If the dripping causes ripples, that should not be hard to fix.

If the surface tension are a problem (is it the resin or the waters surface tension?), why is it a problem? I mean, There is nothing touching the resin, and the water is far below the printing area at the surface.


Can anyone show me these effects on video or pictures? On the available videos, I can not se any waves, ripples or surface tension problems. I can not even on the prints displayed se any errors that I could blame the resin/water characteristics for.

[Feign]

Ripples and waves in the resin were assumed to be a potential problem, but have turned out not to be significant. Surface tension in the top surface of the resin is kind of a more complex problem than originally assumed. Basically, if the laser isn't focused, the resin can develop a skin, like on the top of a cream-based soup as it cools. This sticks to the printed part below the surface and keeps resin from flowing up onto the top of the part correctly. Getting the laser properly focused mostly gets rid of this problem, but it seems to be the biggest remaining source of surface roughness.

The big problems now it seems are that the printer's calibration is still more an art than a science, and the drip system isn't perfectly consistent (close enough for a lot of cases, but not as consistent as say a stepper driven screw like every other z-axis out there.)

[InstantWork]

Good info, thank you!

Can you tell me more about the surface tension problem? For instans if it is chemically possible to lower it? Not a solution to the skinning problem, but it might help anyway...

On the laser focus issue, can you tell me why you have gone with just that specific laser diod/aperture design?

[Feign]

Heh, maybe I should have put that standard disclaimer of mine in the previous post to avoid confusion. I'm not actually part of the Peachy team, I'm just a fan of them who has dome a good bit of research along the way of them making this project happen. I'll answer your questions as best I can, but I don't know any of their internal decision making process.

I can't for the life of me think of an additive that reduces surface tension in oils (or in this case monomers) and if there were one, I can't imagine it having a positive effect on the hardened polymer.

As I said at the beginning, I don't know their selection process, but they did go through several different laser diodes before settling on the one they are using. And the skinning effect I described is at a very small scale and possibly unavoidable without a complex focusing lens setup on the laser that would bring up the price too much.

[rylangrayston]

Feign you did a great job of answering the question!
I love the skin on creamy soup analogy( im going to use that one! )

Ill add a few more things:
1. I Think a small positional offset that occurs when we turn the laser on and off is currently the biggest cause of prints having visible layers
2. physical vibration is probably the second largest cause of layering artifacts
3. getting the exposure per layer in the right range is key in getting really great looking prints.
4. I Think the resin were using is already the lowest viscosity available. warming it up could help lower surface tension even more.

InstantWork To answer your question about our current choice of laser ....

There are many reasons for our latest laser diode choice which is a small laser diode and lens Assembly that plugs into a socket on the board.

1. Because we have designed our own current regulator circuit, we can turn the laser on and off much faster
2 The socket allows you to swap out a laser without purchasing a new board, or full blown laser module.
3. We get major cost savings by purchasing all the laser parts separately, which means we have been able to re-alocate that money to higher quality diode ( i think its from sony) and a better functioning circuit.
4. The newest design is smaller and works well with the idea of making the circuit board part of the physical structure of the peachy printer.

You can see the diode and how it plugs into the circuit in update 36


As for the aperture design it looks like diffraction, ( altho it could still become a problem in certain situations)
Was probably not the cause of holes, our best prints so far(like in update 45) are done with an aperture that was made by pricking a small hole in a small circular piece of a beer can.

[User_Defined]

Feign you did a great job of answering the question!
I love the skin on creamy soup analogy( im going to use that one! )

Ill add a few more things:
1. I Think a small positional offset that occurs when we turn the laser on and off is currently the biggest cause of prints having visible layers

......

1. Because we have designed our own current regulator circuit, we can turn the laser on and off much faster

I have some circuit questions:

Is your Vcc being loaded down from the laser transitions of on/off? Maybe it is pulling too much power at those instants which then lowers the available energy to your amplifiers?

To be clear, in some of your videos, the laser looks like it is moving very slowly, and other it looks to be extremely fast. Wouldn't the slow moving laser avoid the turn on/off problem except for one single point along the shape? Also wouldn't the code moving that on/off location create a spiral (like that spiral you spoke of in another thread)?



If this is all the case, then it sounds like you simply need better decoupling to your laser diode driver circuit. I dont know what kind of current it draws, but an LC lowpass filter should work to limit current spikes.

If it were me designing the circuit and I was facing that problem, I would first measure the spike droop to get an idea of how quickly the laser draws its current.

From there I would take that droop time and come up with whatever the lowest frequency component is. With your droop frequency now found, you can then figure out your filter values.

I would also see about the largest possible ceramic SMD capacitor I could fit in, and parallel that with a smaller 0.1uF.

After that, choose the inductor for your filter which is based on that droop frequency. They have lots of small SMD inductors, or alternatively you could use a resistor and just be less efficient.

For the rest of your problems you're on your own as I am only a hardware guy

hope that helps,




-UD

[mike_biddell]

I have some circuit questions:

Is your Vcc being loaded down from the laser transitions of on/off? Maybe it is pulling too much power at those instants which then lowers the available energy to your amplifiers?

To be clear, in some of your videos, the laser looks like it is moving very slowly, and other it looks to be extremely fast. Wouldn't the slow moving laser avoid the turn on/off problem except for one single point along the shape? Also wouldn't the code moving that on/off location create a spiral (like that spiral you spoke of in another thread)?



If this is all the case, then it sounds like you simply need better decoupling to your laser diode driver circuit. I dont know what kind of current it draws, but an LC lowpass filter should work to limit current spikes.

If it were me designing the circuit and I was facing that problem, I would first measure the spike droop to get an idea of how quickly the laser draws its current.

From there I would take that droop time and come up with whatever the lowest frequency component is. With your droop frequency now found, you can then figure out your filter values.

I would also see about the largest possible ceramic SMD capacitor I could fit in, and parallel that with a smaller 0.1uF.

After that, choose the inductor for your filter which is based on that droop frequency. They have lots of small SMD inductors, or alternatively you could use a resistor and just be less efficient.

For the rest of your problems you're on your own as I am only a hardware guy

hope that helps,




-UD

It certainly sounds like a coupling problem, since the positional error occurs at the point of switching. The implication being that the laser switching glitch changes the current thru the x and/or y coils. As user-defined says.... a simple capacitance de-coupler could solve the problem. So that any energy required by the laser at switch on, comes from the capacitor and not from the rails.

[danl]

...Ill add a few more things:
1. I Think a small positional offset that occurs when we turn the laser on and off is currently the biggest cause of prints having visible layers
2. physical vibration is probably the second largest cause of layering artifacts
3. getting the exposure per layer in the right range is key in getting really great looking prints.
4. I Think the resin were using is already the lowest viscosity available. warming it up could help lower surface tension even more.
...

Some thoughts on above:
#2. Very slight local vibratory effects may come from "burps" from menisci flow at the container and object walls. See comments in #4c below.
#3. Repeatable, exact exposure/layer requires exactly repeated z adjustment. #4c below might be relevant here.
#4 a. Viscosity may not be the only factor...liquid resin yield value and surface/interfacial tension may be bigger players. The use of very low concentrations of select surfactants may help new solid surface re-wetting without affecting solidified resin properties.
b. Previous discussion about waves created by ultrasonics being problematic may be irrelevant if the proper amplitude and frequency is used. Applying the right ultrasonics could overcome any resin yield value effects (eg slapping the side of the ketchup bottle) without creating "waves" that impact deposition/solidification.
c. Water/resin, resin/air, water/sidewall, resin/sidewall, water/print object and resin/print object interfaces and the menisci present at each may be contributing to print artifacts. As the resin and water levels change, the menisci (at water/resin and resin/air interfaces with wall an print object) creep along. That creep, being very slow, may lead to "burp" flow as the liquids "try" to wet the new surfaces they see, but need to build enough energy at the meniscus to overcome interfacial tension.. That is, the meniscus changes geometry slowly without wetting new surface. When new flow finally occurs as the meniscus stretches (eg blow a bubble up til it bursts), the fluid level changes and a vibration occurs. This would lead to z axis "bumps" that might be nearly imperceptible, but possibly introduce layer artifacts. Again, surfactants and/or ultrasonics might address this. Or, having walls coated with a "slip" coat (teflon?) might work.

Hope this is helpful. My motto: "I'm often wrong, but never unsure!"
Cheers,
Dan

[mike_biddell]

It certainly sounds like a coupling problem, since the positional error occurs at the point of switching. The implication being that the laser switching glitch changes the current thru the x and/or y coils. As user-defined says.... a simple capacitance de-coupler could solve the problem. So that any energy required by the laser at switch on, comes from the capacitor and not from the rails.

Oh forgot to say, and at the risk of telling grandma how to suck eggs, the power supply might be slow and unable to deliver energy fast enough, so a Board decoupling capacitor (maybe 100uf) may also help (there probably is one already?).