First experiment

Experiments, experiments, experiments. It's part of the reason for participating in AMRI. So what I've decided to capture data on first is power output of the DLP projector I'm using. In this particular case, I'm using the same projectors as the B9, a Vivitek D535.

For capturing the power output, I have access to a power meter and sensor by Coherent. A FieldMaxII meter and a PM150-50C sensor. Couple of points of interest in experimentation is overall power output of the projector, the drop off in power due to to distance from light source and power density per cm^2. 

FieldMaxII

The first point of interest will be useful when I start testing other projector setups, it'll help form the baseline marker for the various other projectors folks would like to use in a DLP 3D printer. The values gathered will also be good to gauge what power levels are adequate for overall reasonable layer/print times. At the moment, in talking with others working on DLP's in the community, the most important aspect folks look for is lumens (followed by native resolution), most are recommending builders work with at least 2,500 lumens. It'll be nice to see the effect of lumens on power output.

Tested Red projection

Tested Black projection

Tested White projection

One of the question I see being asked a lot is how scalable a DLP 3D printer can be. One would think that pulling back the projector and enlarging to projection area would be all that's needed. Sadly, that's not the case at all, and the reason it'll be interesting to see the power output over various throw distances and size prints. There are a lot of hypothesis and ideas, but it'd be nice to get quantitative values to back things up.

Lastly, power density is interesting to me because it'll allow for us to fine tune our exposure times. The thought is that if we can find the power output over an area, we can use that data to help extrapolate cure times for a specific layer height. On the point of cure times, polymerization rate will also be a future experiment.

Experiments like this is what AMRI is all about, applying the scientific method to create quantitative data for the betterment of research and development. The hope is that having empirical data to back up a lot of things that are taken at just face value will help people make advancements and not stumble over the same problems as others have before them. It's all about anchoring a foundation for which contributors can build upon prior research quickly and adapt results for new ideas.

AMRI and then some...

So I've been pretty busy to say the least, and I apologize for not updating with more content.

I was invited to join the MAKE mag team for this years 3DP Shootout special (should be on news stands in November) and it was a great experience. Lots of great printers and lots of great "printerers" :). That's all I'll say on it as well! Just really glad I got to be involved and a thanks goes out to all that made it possible.

The 3DP Shootout Crew

Spent a couple of days out in San Francisco, CA area and now I'm currently melting in hot and humid Houston. So Houston, I'm currently one of 3 fellows at Rice University for the Advanced Manufacturing Research Institute (AMRI). Linked is Jordan Miller's, he's the awesome mind behind it all, post on the RepRap blog. Check it out and get involved!

I'm here to not only continue development on the DLP 3D printer this blog is built upon, but to also research and develop the same technology for other purposes (related to Jordan's work and tissue engineering) and gather data from various experiments. So other purposes? I did not bring the same DLP as I was building at work with me, so I've got to build a new one while I'm here. Everything is on order and enroute, that's just phase one. After building it out and getting back to where I am back home in the lab, it'll be time to apply some science! I can't do Jordan's work any good by trying to explain it myself, I'll let him post more about that on the AMRI site, but I can elaborate on the experiments and data I hope to collect.

The Gameplan

DLP printers are not something entirely brand new, folks have been working on this for years now. In my recent time with working on DLP printers, there's a lot of questions I've come across and thanks to AMRI, I'll have a chance to maybe answer some of them. Let's start with the peel/tilt/slide mechanism. Curious to find out how much force is needed to separate the recent print from the vat surface? I plan on measuring just that! What about the effects of dyes/pigments on cure times? How about I measure the rate of polymerization between uncolored and colored resin? Done and done! That's just a few things I'm hoping to look into. It'll all depend on time. My main interests lies within various surface coatings.

Aluminum plates for small actuators

Aluminum plates for large actuators 

Sylgard 184 (not all for me thankfully!)

Optically clear FEP tape

Any other thoughts on what I should look into? Now's your chance to ask...so ask! I'm here until at least the 26th of August. Machine should be built out again sometime next week.

Until next time, remember to keep your goggles on!

SAFETY FIRST!

Printing Updates

So the two following photos pretty much sums up how things are going right now.

Vat soaking after a few "prints"

Small portion of the trials...

So it's kind of been a failure. Resin is curing, but I'm still fighting some issue. Luckily it isn't the more difficult of problems (that being peeling of the print off the vat bottom, that's working out just fine!).

Problem #1 is over contamination of UV. I believe this is due to the fact that I have a transparent/clear build base/platform. The light is being absorbed and then refracted/reflected/etc. to the surrounding areas. 

My current solution is to make it out of something opaque. You can see that the current platform in the 2nd picture has spots of black. That was my attempt at making the base opaque with spray paint. Which in itself lead to another discovery, the resin and spray paint don't play too well with each other!It pretty much came off rather quickly.

Problem #2 is print adhesion, I can't quite get the print to stick to the build platform. I'm sure adding spray paint didn't help that situation either.

Solution to that issue is to experiment with various different platform/base materials. Having exposed the resin to bare acrylic, I know that that is a not a route I want to go...well unless I want to use a chisel to remove my prints. It sticks REALLY, REALLY well. Seeing what TristramBudel (that dude with the awesome Instructables! go vote for him so he can win that laser!)is using, I'll try and sample various materials he suggests and see what that yields.

And on that note, after seeing his instructables, version 2 is def going to be a horizontal projector. Makes things WAY more compact. I like it a lot and its given me lots of ideas. 

First Run

So, let's run through the setup I have going heading into the first run. I'm running RAMPS 1.4, loaded with Marlin (full stepping on all axes, 200 steps per for the THK KR20's). I'm using Gary Hodgson's flavor of PrintRun with ProjectLayer. To slice the STLs I'm running Slic3r 0.9.10b, using a layer height of 0.1mm. I'm not running a tilt mechanism at the moment, mainly because I'm hoping to get away without one...but we'll see how that pans out.

And the first object to be printed is...the Goldberg Polyhedron! A little crazy for a first print, but to hell with it! I was skeptical of the model in its current state, it is a sphere after all and there is minimal surface to make solid contact with the build platform. Not likely candidate for success.

Going back and trying it again next week, hopefully things will get better. The new plan is to create a raft/support structure to help with adhesion, will likely have to model that into the STL myself.

Vat Coating #1: Sylgard 184

So, new vat is constructed. Nice and water tight! Life is easy when you have access to all the tools you need isn't it?

Now that the vat is constructed, onto the coating. This coating is whats going to help prevent your cured resin to adhering with the bottom surface of the vat, we don't want that because we want to 1) be able to lift the platform to do the next layer 2) lift/tilt to allow resin to flow back into the just cured area. I have purchased Sylgard 184 and PTFE film (adhesive backed and non-adhesive backed), check the hardware section for links) for this task. 

First up is Sylgard 184. This seems to be where most folks with bottom-up resin printers start with. It's what the B9 uses and can be found on the various DLP printers you can see on Youtube. In general, this stuff is used for solar panel constuction...it's silicone that is non-UV retardant and cures clear (pretty much optically clear). Makes sense for what it's used for. Make sure if you do buy some Sylgard 184 that you get the legitimate stuff that is Dow Corning brand, there are knock-offs of.

I will say that a lot of the folks I talk to advise on not using it, but me being the stubborn person that I am, will give it a try anyways.

Sylgard 184 is a two part solution that needs to be mixed with the ration 10:1. The vat I currently have has an interior area of 5" by 6" (pretty much an 1" offset of the build platform I have, which is 3" by 4"). Doing some basic calculations, I arrived that I would need approximately 45-50 mL to get a decent coverage. Whipped out a small beaker and got to mixing.

During the process, the two part mix will get really smoky and full of bubbles. You might think this is a bad thing, but rest assured, all those bubbles do indeed disappear. Mix for 3 or so minutes, you really want to make sure that things are mixed really well for best results (and of course make sure you measure accurately for the 10:1 ratio).

Find a nice and leveled surface for your vat (a leveled surface is key in getting an even bottom for your print surface), pour in the Sylgard 184 and spread it evenly. Gravity will take care of the rest and things will settle. Now leave it there for 24 hours (probably more time then needed, but it can't hurt to leave it a bit longer!). On that note, you can cure the Sylgard faster by placing it in an oven up to 80C (careful, acrylic may warp, this might be a good reason in the future to use glass). If interested in that route, be sure to read the info on the Sylgard 184. 

Few minutes after...

about 30-45 minutes after...SO CLEAR! I did use a need to poke a few air bubbles.

Resin should be in tomorrow (I guess that's now today at the time of this writing) and I should have the vat and baseboard should be constructed then too. Should be "printing" tomorrow...printing well or not is a whole other story. Exciting nonetheless!

Build Base and Vat

It's where all the action happens, your build base/platform and the vat. The vat is what contains your resin/material and the build base/platform is where the print adheres to.

I decided to construct my own base and vat. It was fairly straightforward for constructing both with the help of some Weld-on #4. 

For the vat, I laser cut all the pieces, out of 1/4" acrylic, I needed as well as a jog, popped the pieces in place and dripped the solution into the seems. Capillary action takes care of the rest, one drop is all it needs!

Holes on the corners for bolts.

This little slot is an idea for the tilt mechanism a friend thought up...

The base assembly was also entirely laser cut  I have access to a 120W Trotec laser cutter. The higher wattage is great for cutting/engraving through thick materials (1/2 or 12mm+). I used 0.5" acrylic as the build medium. The thought was to layer the pieces together, like so:

Side View, see the depth of the engraving for nut trap

Overhead view

Partially assembled, still need the last piece with the glass epoxied together.

I will be posting all of my laser cut files as soon as I'm sure they are reliable. If you want them sooner, shoot me an email.

The Setup/Partial Assembly

From here on out, the BoM will be moved over to the hardware section. You will be able to find snapshots of carts from the various vendors in which I acquired parts from. Same goes for the software setup I have, that includes electronics, FW, slicer and host.

Also on the note of BoM's and supplies. I picked up a fair amount of the linear actuators for really cheap. If you are interested in some, please contact me and I will gladly sell you some. Drop me a line via email or state your interest in the comments.

THK KR20 actuators

Now to the stuff you really care about...assembly!

It's a more or less straightforward build, I'm sure the pictures will speak for themselves. Do let me know if any part could be bettered with a caption.

The workflow that I followed for the frame assembly was the following:

Base -> Front/Back -> Projector Mount -> Left/Right -> Tilt axis -> Top -> Z axis

 

More photos to come shortly... :). I'm a little behind on posting since the last update...been busy curing resin! :P

All the while, I'm thinking of different methods for executing the vat tilt mechanics. So many options. Different methods for tilting to different methods for surface coatings.

If you're interested in learning more about DLP 3D printers and want a more immediate response, consider joining fellow contributors on IRC via the #dlp3dprinting channel on Freenode. You can find me on their under the handle "digiFAB."