Infill layer discovery.

I just remember I have a blog. Is a bit dusty but why not give it a try and give something back to the internet. This is mainly about 3D-printing.

So I have dipped my toes in a new hobby, Quadcopter flying. This mean some crashing which mean my 3D-printer will have to do some work. I made some alteration to some legs I found on thingivers.com and here it is in all is glory.


I few days later and this is what it looked like. I hit a tree and then the quad hit the ground. I thing the wind I my inexperience had some thing to do with it.

Anyway. A new arm take some time to get shipped and I do have the RepRapPro Mendel that should be able to do the job. I haven't been using my printer that much lately because of lack of time and relay no need and then there is this insane calibrations and testing that has to be done before a print succeed. It might be just me but I never felt I'v master it. During my attempt to print a new arm I found some thing I have not realised before and I thought I share it. It might be common knowledge but if so I have missed it out there.

The printing of the arm.

I downloaded a new arm from thingivers.com, http://www.thingiverse.com/thing:181301. With not much calibration I manage to print one. Honeycomb infill pattern and maybe 50% infill, I don't remember.

However it had some strange things going on. I imported the .stl in sketchup and cleaned it up and beefed up some parts I thought was a bit to thin. If it hod during flight I publishing it on thingivers.com.

With the honeycomb pattern for infill I figured I could get away with only 10% infill so I change to that.

I started a new build and after making sure it started ok I went to bed. This is what I found next day.

I haven't figured out exactly why it failed but I have an idea.

Printer upgrade to come.

During this time I also printed some new parts for the printer to upgrade the x-axis to the new design of the RepRapPro Mendel. They came out ok but do have some holes in the top and some other imperfection. I think they've will do so I have no plan redoing them.

At one time during this print one of the bearings on the x-axis on the printer come loose. I think it is related to the failure of the second attempt to print the quad arm but I didn't know that then.

Callibration

I thought I make an effort and calibrate the printer some more and specially the bride capability. http://www.thingiverse.com/thing:269679. I never rely made bridging work before and when trying now it was not pretty.

After some testing with the speed and the flow I had something I could live with. It is not perfect and wont win any contests but it will do for the kind of stuff I see me doing. What I ended up with as for settings is a low bridge speed of 20 mm/2 and a bridge flow ratio of 1.2.
Back to the arm.

Two more tests and 2 more failure.

The first one I didn't see, that's the upper one, but the bearing was off again. The second I say when it started doing the infill layer before the top and it was a disaster. What happen was that the layer did't lay down smoothly but curled up and made high spot over the howl area. Since the area is big they cool down and become hard. Then when the hotend was speeding back to the other side along the whole arm to do some printing it hit those spots and that was a bumpy ride. It made the print go of alignment and I guess this is also what made the bearing come loose. I stooped the second print asap.

Some thinking and searching I formed an idea about what was wrong. With low infill, in my case 10%, laying down the first infill layer is like birding but is not treated like that and I had not calibrated that. I use Slis3r and there are different speeds for the two. With that in mind I did some more tests.

Test test test.

Here the first one is with original settings and the second has infill layer settings close to my bridge speed settings. It's not much of a different but the second is a bit better. I figured it has to do with box being to small to show the effect of low infill.


 

I made a bigger one and stopped the print after it started the second layer.As can bee seen the first infill layer is much better then on the two failed quad arms.
A new arm.

Time to test my new found knowledge on the arm. Success!

I did change the infill to 15% so it probably helped some to.

I realised I don't need the small leg since I'going to make some smaller ones to put under the hub of the quad. Like the ones found here, http://www.thingiverse.com/thing:34562.

Some kind of conclusion:

There could be more tests done to rely nail this down but I leave it for now. This is what I think is going on.

When having a low infill the first infill layer starts looking like a bridge and the settings need to be calibrated much better.

With high infill the quality of the first infill layer seams not to matter as much probably because the next layer will compensate and hide problems to some extent.

I don't know what would have happened if the hotend had not taken the path over the high spots. It might have heated them and smooth them out on its next layer.

I guess that depening on how much the highspots cools down also plays a role in how well the second layer can hide it.

Looking at my first print one can wonder why it mad it through the whole print but I think it has to do with the high infill making the honeycomb pattern small which makes it much easer to get the layer flat.

I hope someone will get some help from this.