Bridging issue - assistance requested

[drayson 75]

Dear all,

today I started some bridging tests - unortunaltely with very low sucess.

Therefore I´d like to ask for some hints/assistance on this as I need to improve my printing quality for a request I got...

I can remember that - I think Illuminarti?? - had really great results and I followed the posts but honestly I have no idea what I should try further...

Here two pictures - top looks fine, but don´t turn it upside down... :-(

This part was printed at UM1 with Cura14.01 and following setup:

layer: 0.1

shell/bottom/top: 0.8

fill: 50%

speed: 60

temp: 200

fan: 80%

Also tried with temperature set to 195, 205 and 210 with same/less success

I would appreciate your assistance

Thanks in advance,

chris aka drayson

 

[Dim3nsioneer 557]

Did you see http://umforum.ultimaker.com/index.php?/topic/3406-experiments-on-bridging/?

From my tests I think yours are definitively more on the sunny side already... :wink:

What material did you use?

There might be some tricks like changing the acceleration for the bridge layer to avoid single strings from ripping but the sticking together of multiple strings is very difficult to avoid.

Maybe foehnsturm can say something about the effect of the crossflow fan onto bridging? Any difference concerning the sticking together?

EDIT: If you printed it oriented as shown in the first picture (x direction) and if you use the stock fan or any other fan duct blowing from the left only try to rotate it by 90° and print it in y direction.

 

[mastory 44]

Also running UM1 here and desire to improve my bridging performance.

It seems to be somewhat of a black art. Is there definitive "guide to bridging" or some such thread or reference that could explain the in's and outs of the prevalent theory?

I have read lately that controlling acceleration (slower) on the bridge layer can make a big improvement. Is this be controlled separately than acceleration used on infill, loops or perimeter? Settings in Marlin I'm guessing

Is it possible to control cooling differently while bridging?

Thanks again.

Drayson,

There is a big difference in the print properties of one filament to another and from manufacturer to another. In my experience (may be anecdotal) some colors work better than others. For me, I am getting better prints from the lighter pigments. Grays and blacks are my least liked. I have to admit, I can't support this opinion with imperical evidence. I just haven't done enough printing.

Matt

 

[drayson 75]

Hi !!

Sorry, seems that I missed this one - thanks for pointing...

Material is UM PLA-filament, grey/silver, the fan I use blows from both sides - my own design.

()

I will try TweakAtZ but I don´t think that on a much more complex part with several bridging sections it might working...

One problem I additionally have is, that the surface/first layer above support structures is a mess - that´s the initial root for my bridge testing.

 

[gr5 2,173]

The link that dimensioneer posted above is the best information on it. If you look carefully at the magnified images you can see that some bridges break on that first layer because of underextrusion for the first few millimeters so acceleration helps a lot. If it were me I would edit the gcode by hand for experiments. An acceleration of 1000mm/sec/sec is much slower than typical (3000 or 5000) but is still kick ass fast. I think non-ultimakers are around 100-200mm/sec/sec. Like the makerbot but not sure.

The gcode command is:

M201 X500 Y500

and then restore with

M201 X3000 Y3000

I also did a bridging test once - not as detailed - but here it is (post #17):

http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=25304

 

[mastory 44]

If there is a pronounced difference, shouldn't the slicer adjust the accel before and after a bridge layer automatically. As well, bridging temp and cooling preferences. May not be possible with the temp due to the lag time for a temp change.

 

[gr5 2,173]

Of course but right now this is still experimental. Someone needs to prove it really works and do lots of tests and then if it really works, someone should next write a plugin that maybe locates these bridges (might be hard if there isn't a comment in the gcode although cura puts lots of helpful comments in there for plugins to use).

Then when it has been tested thoroughly by at least 3 people, and Daid has tried it and is convinced it works, only then does it make sense to add to Cura.

First we need to see if it helps at all. I am kind of doubtful. Actually I think what would make more sense is to extrude a little bit extra just before the start to make it a tiny bit thicker - maybe pause for 1/10 second before starting the bridge - I'm not sure. You could split the single bridge move into two moves - the first move would be over extruded and only move 1mm or so and slow - maybe 5mm/sec, then do the remaining move with normal extrusion.

 

[mastory 44]

Fair enough. If I can get any reasonable results, I will try to do a small test this week.

As an aside, I see that Slic3r has separate speed and accel settings for bridging. Never used it so only a casual note. Maybe I'll give it a try too.

 

[drayson 75]

Great to see that there are a lot peaple having the same interest in bridging as me... :-)

@ Matt: I already recognized too that it is - as everything - strongly depending on the filament.

My concern with a plugin is that e.g. with TweakAtZ you have to identify the bridging layer manually. With a more complex structure, this is a bit of work that I don´t like to have in operational work.

I´m not a software guy (ironically, I work as a manager in a software dept.), so I have no clue about generating code, but I would also assist in testing....