increasing print speeds

[helifrek]

Hey guys,
Another question to bug you experts with..... I have searched the forums for speeding up my prints and read a few posts but I would like to go into a little more detail. Right now I am getting great quality prints, I do get a bit of a seam on my prints which I am slowing adjusting but overall I am happy with the quality. Well, I have been getting more and more orders for larger pieces and I work a full time job, I don't like leaving the printer running while I am away at work and some of these prints take 10-15 hours to do! besides just using the speed toggle in the machine control panel to speed up my machine, how do I speed up the overall print speed. I am using Rsilvers PLA medium profile and the default printing speed is set to 4800 mm/min. Should I just slowly increase this number until I see a decrease in quality and then back it off a bit? What are some other things to look out for when speeding up the printer? I have watched some of those high speed 3D printers in action and they are impressive but I also don't want to kill my M2!

Any advice is greatly appreciated!
Brandon

[Matt_Sharkey]

I run my Infill at 120 mm/s (7200 mm/min) and it seems to do fine. I would only up the speed on infill really because you depend on the perimeters to show quality.

Other ways to expedite the process is in other parameters, with PLA you can do 15% honeycomb infill and it turns out pretty good. also since PLA can push out huge quantities, you could up your layer heights to 0.30mm, just make sure your extrusion width is 0.55 or more.

I haven't tried printing with large layer thicknesses and widths, but I imagine your dimensional accuracy will be greatly diminished

[ednisley]

more and more orders for larger pieces

Perhaps you should ask "How much faster must my machine operate to make a difference to my business?" The answers will put bounds on the effort you're willing to put into it.

If you must increase production by a factor of two (or more), then you need another 3D printer (or more), because you flat-out can't melt plastic and slap down the thread twice as fast (or more) with anywhere near the same quality. That would reduce a 10 hour session to 5 hours, while doubling your capital. If you're charging enough to justify your time and materials while amortizing the printer, that might be an easy decision. Conversely, if you're underpricing your work, then you're running a charity.

You might get an additional 50% out of your existing setup after balancing neurotic speed tweaking against quality degradation. That would reduce a 10 hour session to 7 hours, at the cost of your time and considerable scrap along the way.

A 10% increase will be straightforward and gives you a 9 hour session instead of 10. In a real factory, production managers will perform unnatural acts to increase production by that much, but maybe you should just back off your morning dose a little... [grin]

Increasing infill speed while maintaining the same outer perimeter & surface speeds will certainly speed up the process, but you'll soon discover that the outer shell soaks up nearly all the print time and you can't increase that pace by very much at all. That's particularly true for parts with a high surface-to-volume ratio: anything small or plate-shaped won't benefit from faster infill.

Testing your slicing changes with a G-Code previewer (or, better, jsc's simulator) will give you a rough estimate of how much improvement to expect, so you need not waste plastic on "improvements" that don't matter.

But it's great to have customers who want what you do, isn't it?

[helifrek]

Thanks for the advice guys! I guess I thought my machine just seemed slower than a lot of the ones I have seen on youtube and that maybe my settings were a little off. Right now I am printing 8 pieces that take up my entire bed, S3D estimates the time will be about 7 hours but in practice I know it will be closer to 9 or 10 hours, I was hoping to increase the speed enough to maybe get closer to the estimated time.
In all honesty it isn't really the machine slowing me down, it is the fact that I leave for work around 6am and I don't get home until 5pm, so an 11 hour work day plus 7 hours of sleep comes to 18 hours, add cleaning, shopping, getting stuff ready for the next day, time with the wife...... it leaves me with very little time to babysit the printer and I don't trust printing while I am at work, the extra speed would at least let me print through the night and finish before work on some jobs.

I have a year left at my current job and then I am out! I am going to start my own business doing 3D printing, laser cutting, vinyl cutting and whatever else I can get into. When that time comes I won't have to worry about the speed so much and I cannot wait!

You are right, it is nice to have return customers that are impressed with your work! I will eventually be ordering a couple more printers to keep up, I wouldn't mind something with a larger build area and an SLA printer to mix things up a bit. The parts I am printing now are 18" long and had to be split into two 9" pieces! I wish makergear would make a 24x12x12 sized printer with the same quality as the M2 hint hint......

Thanks again guys, I will try and increase the infill speed. Can't I also increase the infill layer height while keeping the perimeter at .2?

[Jules]

You'll want the infill height to stay the same as the layer height, but you can really see a significant improvement in speed by printing with a 0.25 mm layer height instead of 0.2 mm. (Also want to change the layer width to about 0.45 or so to go with it.) I haven't noticed any reductions in quality by going to 0.25 mm.

And print with a single outline. Cuts the time significantly.

It adds up.

[helifrek]

You know, I never really considered increasing the layer height to .25.... I am so used to .20 that I never gave it a second thought. just doing that cuts over 2 hours off of my estimated print time in S3D! also went to .45 extrusion width as suggested. Doing some more thin walled calibration squares before I restart this print, I think I had my overlap turned up too high because on the inside of the perimeter walls it was forming bumps where the infill lines met..... Also it is oozing a little more than usual... going to drop them temp a little and increase retraction. fingers crossed, this is a huge order and I need to have it shipped by Monday!

[helifrek]

Another question while I have you guys here, do any of you have problems with a warped piece of glass? I use a dial indicator to level my be and I can get the entire piece bed perfect except one side...... any ideas on how to fix this?

[Matt_Sharkey]

nothing is perfect. most likely the discrepency comes from the X gantry being bowed down by the belt tension. you could try loosing the tension a bit (and slow down your travel speeds because the belt is more prone to skipping now.)

you might find that satisfies you, otherwise, go for Mic6 tooling plate.

but it's worth saying, nothing is perfect.

[3dPrintingMD]

Another question while I have you guys here, do any of you have problems with a warped piece of glass? I use a dial indicator to level my be and I can get the entire piece bed perfect except one side...... any ideas on how to fix this?

What I found is that there is also some deviation in the rubber bed corners. At the end of the day the glass sits on the heated bed, and that in turn sits on those rubber bed corners. The thickness of them is different if you remove and measure each one. That will haunt you in trying to level. If the glass is warped, then flip it over and see your results. A bend one way is a bend in the opposite direction when flipped, so you will quickly be able to see if its the plate.

At this point I've worked to take the variables out the system the best I can (so far). PETG Bed Corners, each to the exact size, I printed a bunch of these and got 4 that measured the same thickness. I went with an MIC 6 Plate. Super flat.
And that's where I stand now. The only other issue in the system, and I'm not willing to fix, is the thickness of the heated bed, the actually heater part could deviate from corner to corner.

[helifrek]

Thanks for the replies, the .25 layer height and .45 extrusion width did help a lot! I would like to mess around with turning up the print speed but I don't feel I know enough to be able to do it correctly.

Back to bed leveling, I put my dial indicator back on and cleaned everything really good. I messed with it for about 30min getting the level as close as I possibly could but there are still deviations that seem to be in the glass. There seems to be a bow (or dip if I flip it over) in the glass on the side I usually put facing the front side right in the middle. I can even see on some of my large prints where there are gaps in the lines on the first layer even after I have leveled the bed and done an extrusion and Z axis calibration. It has me pulling my hair out! I increased the first layer to 125% to help a little. I will look into the tooling plate next.

Thanks again guys