Advanced discussion on Extruder Temp drop and Warping

Zortrax M200 General Discussion
#1

Hi guys,

First of all - long time user, long time reader of the forum, own 2x M200's (one with Z-temp), so please don't take my post as one of a noob :)

NOTE:

I'm printing using Z-ABS

0.29mm Res

Normal Speed

Light Infill

Angle Support: 20 (lite)

Fan Speed: tried all, though 0% works best for this part.

Surface Layers Top/Bottom: 5

Seam Normal

Smart Bridges Enabled

V2 Perf Plate

V2 Hotend

I've been having trouble recently with my M200 printing large parts. I'd like to talk about 2 issues, which are probably connected.

1. The temporary Extruder Temperature drop after printing the raft.

2. The Warping of parts.

To use an example I thought I'd post one of the worst cases I've seen. Please see attached STL "Rig 2". This Flange takes up most of the print bed and I've never been able to print it without it warping and having horrible looking surfaces.

Issue 1:

When the raft is finished being layed down (perfectly every time btw), there is a temporary Temperature drop in the extruder. I'm assuming this is to make the part easily separable from the raft when the print is done. In other words, the extruder wants to heat the first 'part' layer just enough to stick to the raft and no more. Whereas every other layer of the print is nicely bonded together at the appropriate temperature (~275 Degrees C for Z-ABS).

       Now please look at photo 1. You can see the stringing effect that this causes. The extruder has cooled down temporarily, and the first 'part' layer has started to be layed down. However the temperature is not high enough to sufficiently 'bond' this layer to the raft. Each pass of the extruder lays down a line which is literally free to move. Some of these lines can cross over each other as well which causes every subsequent layer to be uneven. If the print was to keep going, the end result will look like Photo 2. As you can see the top layer is terribly uneven. I've seen plenty of posts about stringing on here and how to solve it, but little about what causes it in the first place!

How would any of you combat this issue for a part like this? Please feel free to print the part and see how you get on with this issue.

Issue 2:

This is a oldy but a goody! I'm assuming Zortrax has realised how bad the M200 is at utilizing the entire bed for large parts - hence inventure's ridiculously small print bed! Anyway - I've invested this much time into the M200 I'm determined to make it work. For this part I've obviously calibrated the bed just prior to printing, put doors on the sides, got the room temp at a constant (~22 degrees C). By the way I'm convinced the room temp is a huge factor in warping - I know some of you 'advanced user's' aren't convinced ;). Sometimes I use glue stick but since investing in the V2 perf plate, I haven't bothered. (I should point out that most of the parts I print are small and don't need it). Also, I haven't used glue stick in the Rig 2 part - The vast majority of the warping I've seen on this part is due to the aforementioned 'extruder temp drop' issue above. So the part warping off the raft as opposed to raft warping off the bed.

      See Photo 3. Rather than bitch about this result - I thought I'd suggest something constructive for M200 users. I propose we make a check-list of variables which every user can tick off to try and combat the warping issue.

What do you guys think?

Lastly, if anyone can print the part above successfully then please let me in on the secret. Thanks for reading :)

#2

Hi guys,

First of all - long time user, long time reader of the forum, own 2x M200's (one with Z-temp), so please don't take my post as one of a noob :)

NOTE:

I'm printing using Z-ABS

0.29mm Res

Normal Speed

Light Infill

Angle Support: 20 (lite)

Fan Speed: tried all, though 0% works best for this part.

Surface Layers Top/Bottom: 5

Seam Normal

Smart Bridges Enabled

V2 Perf Plate

V2 Hotend

I've been having trouble recently with my M200 printing large parts. I'd like to talk about 2 issues, which are probably connected.

1. The temporary Extruder Temperature drop after printing the raft.

2. The Warping of parts.

To use an example I thought I'd post one of the worst cases I've seen. Please see attached STL "Rig 2". This Flange takes up most of the print bed and I've never been able to print it without it warping and having horrible looking surfaces.

Issue 1:

When the raft is finished being layed down (perfectly every time btw), there is a temporary Temperature drop in the extruder. I'm assuming this is to make the part easily separable from the raft when the print is done. In other words, the extruder wants to heat the first 'part' layer just enough to stick to the raft and no more. Whereas every other layer of the print is nicely bonded together at the appropriate temperature (~275 Degrees C for Z-ABS).

       Now please look at photo 1. You can see the stringing effect that this causes. The extruder has cooled down temporarily, and the first 'part' layer has started to be layed down. However the temperature is not high enough to sufficiently 'bond' this layer to the raft. Each pass of the extruder lays down a line which is literally free to move. Some of these lines can cross over each other as well which causes every subsequent layer to be uneven. If the print was to keep going, the end result will look like Photo 2. As you can see the top layer is terribly uneven. I've seen plenty of posts about stringing on here and how to solve it, but little about what causes it in the first place!

How would any of you combat this issue for a part like this? Please feel free to print the part and see how you get on with this issue.

Issue 2:

This is a oldy but a goody! I'm assuming Zortrax has realised how bad the M200 is at utilizing the entire bed for large parts - hence inventure's ridiculously small print bed! Anyway - I've invested this much time into the M200 I'm determined to make it work. For this part I've obviously calibrated the bed just prior to printing, put doors on the sides, got the room temp at a constant (~22 degrees C). By the way I'm convinced the room temp is a huge factor in warping - I know some of you 'advanced user's' aren't convinced ;). Sometimes I use glue stick but since investing in the V2 perf plate, I haven't bothered. (I should point out that most of the parts I print are small and don't need it). Also, I haven't used glue stick in the Rig 2 part - The vast majority of the warping I've seen on this part is due to the aforementioned 'extruder temp drop' issue above. So the part warping off the raft as opposed to raft warping off the bed.

      See Photo 3. Rather than bitch about this result - I thought I'd suggest something constructive for M200 users. I propose we make a check-list of variables which every user can tick off to try and combat the warping issue.

What do you guys think?

Lastly, if anyone can print the part above successfully then please let me in on the secret. Thanks for reading :)

Issue #1:

Ribbon cable / Heater&Thermocoupler / PCB

Issue #2:

Its most likely ABS shrinkage, I can make you checklist here:

  • Use filament with less shrinkage factor than ABS (Z-ULTRAT / Z-HIPS / Z-GLASS / Z-PETG)
  • Use side covers
  • Make sure that temperature is stable and avoid air flows
  • Use ABS Juice or any 3D print glue

For the bad top layer I recommend you printing it with 0,19 layer. I will make test with 0,19 and 0,29 so you can compare.

#3

Phil,

Correct me if I'm wrong but In my opinion there is only one way to print this part without any problems.

And that's on his side.

It's certainly not the most economical way, because you have to print a lot of support material.

Still the round shape will have ugly corrugations if you print it with a layer thickness above 0.19.

And the holes will have strange forms (sometimes)

But you will get the best visual results.

About the temp. drop and warping causes...

It's a good Idea really - but I think that Zortrax should solve this in the first place.

from what I can read here on this forum there are enough users with the same problem.

and that's mostly warping of large parts.

They can say whatever they want but I have not seen one large "engineered" part on this forum

that's perfect.

So a better question would be bring up the proof and tell how you did it !

But that will be a problem for a lot of people lurking here. Because they have to admit they don't know how to do it.

And a few who do know how to do it wont tell you how to solve it because they have a (understandable) economical advantage now.

So my conclusion in all for the M200. Is this printer really a so called Prosumer 3D printer. Yes in some cases.

and that's why I strongly believe that the inventure is proof of this. Think about the smaller printarea etc...

On the other hand I think Zortrax should do more to improve the M200 like the promised Wifi connection to say something

Because without the support (read money) of the M200 users they would never be able to build the inventure.

#4

First of all - many thanks Marcin and Dirk for the responses!

I was going to create a new thread again for this as I've done some very interesting analysis of the Extruder Temp Drop issue :) However I'll just keep this thread going for now.

Marcin - I have checked and re-checked connections for Ribbon Cable/Heater & Thermocouple/PCB. Even put new grub screws in for the extruder/Heater & Thermocouple.

Dirk - I completely agree and appreciate the honesty! You're right - there's no point in Zortrax fooling us or vice-versa. This thread had 30/40 views before I got a reply. I bet everyone took one look at the part and thought....no!  :D

To echo Dirk's comments - I would challenge anyone reading this to print the part above! ;)  

Unfortunately it's too large to be printed on its side. But that seems like a cheat to me anyway.

Onto the interesting bit....

So I've installed Z-Temp on my second Zortrax (this printer has the exact same issues described above). I decided to log the Voltage of the extruder over the first 7 layers, convert the voltage to temperature, and voila!

Please see the 2 screens of my spreadsheets attached - both prints fail when the part layers start to print. The interesting bit is the graph showing the Z-temp offset at -50. This one fails at 215degreesC. However the extruder can print PLA at 215 no problem - I proved that on the -60 graph!!

What's going on here?!! Losing the will....

Again if anyone wants to clarify what SHOULD be happening when the Extruder drops temp while printing the part layer....please let me know :)

Thanks again for reading!

:)

#5

Phil,

Thank you for sharing this info with us users here !

I think this should be enough proof if I may call it this way to address a wake up call to Zortrax. But believe me they already know it.

And I think they need help from us users to solve this problem.

I can only say it's not necessarily a bad thing to admit you have a problem.

But it will become one if you neglect the knowledge/experience of the user.

Let me say this I'm certainly no electronics specialist, but I do know as a mechanical specialist when something electric is wrong.

#6

First of all - many thanks Marcin and Dirk for the responses!

I was going to create a new thread again for this as I've done some very interesting analysis of the Extruder Temp Drop issue :) However I'll just keep this thread going for now.

Marcin - I have checked and re-checked connections for Ribbon Cable/Heater & Thermocouple/PCB. Even put new grub screws in for the extruder/Heater & Thermocouple.

Dirk - I completely agree and appreciate the honesty! You're right - there's no point in Zortrax fooling us or vice-versa. This thread had 30/40 views before I got a reply. I bet everyone took one look at the part and thought....no!  :D

To echo Dirk's comments - I would challenge anyone reading this to print the part above! ;)  

Unfortunately it's too large to be printed on its side. But that seems like a cheat to me anyway.

Onto the interesting bit....

So I've installed Z-Temp on my second Zortrax (this printer has the exact same issues described above). I decided to log the Voltage of the extruder over the first 7 layers, convert the voltage to temperature, and voila!

Please see the 2 screens of my spreadsheets attached - both prints fail when the part layers start to print. The interesting bit is the graph showing the Z-temp offset at -50. This one fails at 215degreesC. However the extruder can print PLA at 215 no problem - I proved that on the -60 graph!!

What's going on here?!! Losing the will....

Again if anyone wants to clarify what SHOULD be happening when the Extruder drops temp while printing the part layer....please let me know :)

Thanks again for reading!

:)

Great job but I've stopped reading it after Z-Temp and PLA. 

#7

Great job but I've stopped reading it after Z-Temp and PLA. 

You should look up Zeepro and see how they are doing.

#8

Again if anyone wants to clarify what SHOULD be happening when the Extruder drops temp while printing the part layer....please let me know 

I think this behavior is pretty normal and expected. Pretty much any extruder on any printer is going to drop temperature if you blow room-temperature air on it. The recovery time could conceivably be shortened by delivering more power to the heater, but then the PID loop would become touchier.

Some of the blame may go to the fan duct design. I raised this issue on another forum recently, and one of the "experts" there said "Well, MY fan duct design doesn't cause any temperature drop at all". I'm skeptical, but haven't tested his duct. Visually/intuitively, the Zortrax duct looks good to me...

With ABS the temperature drop is not a big deal, and in fact I think Zortrax takes advantage of it to make raft removal easier.

The temperature drop can be roughly halved by wrapping the block in flexible ceramic insulation. Using 100% or lower fan speed rather than Auto will also reduce the fan shock, but then you lose the advantages of the intelligent fan speed control.

#9

You should look up Zeepro and see how they are doing.

Zeepro ?

Oh yes , the guys that sells a printer wich work with filament cartrides at the cost of 100$ per Kg :)

#10

Phil, I guess that the main problem with your print is that is an almost impossible part to print in one shot  because of his mass, over 100 grams and his size.

The only intuition I got , that might be a total mistake or a brilliant solution, is to insert in the print process some pauses to let the printed material to reach an even temperature, then continue the print.

I agree with you about room temperature , the higher the better , I finally put my M200 on the top my pellet boiler the warmer place in the house…

#11

Rather than bitch about this result - I thought I'd suggest something constructive for M200 users. I propose we make a check-list of variables which every user can tick off to try and combat the warping issue.

What do you guys think?

Lastly, if anyone can print the part above successfully then please let me in on the secret. Thanks for reading :)

I havent experimented enough with the printers yet, but I'm suspecting it will be very difficult to get warp free print on many parts without a heated build chamber. I printed some compliant mechanisms a couple weeks ago and when I took my heat gun to finish the parts, I could cycle (reversibly deform) the components using only the heat gun by several mm to cm. I suspect the same thing happens in the printer.

#12

Hi guys I've had some success - so i'll report my findings below, and also respond to some comments in kind.

Thanks to everyone for your help though - any comment is much appreciated!  :)

First of all - printing at 0.19mm seems to be the optimal for large parts. (thanks to Marcin who suggested this in his first response above^). I've literally tried every other layer resolution and I've stopped the prints less than 1/3 the way through as I can already tell how bad the part will turn out. I would say this is a firm conclusion here - Zortrax may want to verify or take note. This info would be very handy for beginners/intermediate users as part of some sort of 'how to guide' literature. In other words, for a part of size LxWxH or greater, a layer res of 0.19mm is recommended.

Great job but I've stopped reading it after Z-Temp and PLA. 

I respect your opinion Marcin as I'm assuming you're a Zortrax purist, however I'm happy to report that Z-temp has unlocked my M200 to its full potential, and hopefully I'm getting less cancer as a result of not breathing in the styrene fumes of ABS  :D. Personally I'd love to see a section on the forum for third party hacks!

Phil, I guess that the main problem with your print is that is an almost impossible part to print in one shot  because of his mass, over 100 grams and his size.

The only intuition I got , that might be a total mistake or a brilliant solution, is to insert in the print process some pauses to let the printed material to reach an even temperature, then continue the print.

I agree with you about room temperature , the higher the better , I finally put my M200 on the top my pellet boiler the warmer place in the house…

That's kind of what I'm trying to achieve though lucamac. I'm trying to prove/disprove that you can utilize the entire bed for a large part, and if me/or anyone can do it then create a checklist of settings which we can all repeat  :) . I like your suggestion of inserting pauses to allow the material to settle. I haven't tried it, but as you can see below, I didn't need to  ;) .

Anywho, onto the results!

I printed 2 very similar parts to the one in my first post. So ok - it's not the same, but as you can see they are large parts (155mm OD) and trust me they will give you warping issues  :lol: .

I suppose most of you will be interested in Part 1 (red one), as it is Z-ABS, and printed on a non-modified M200. Here are my settings:

Z-ABS

0.19mm

0% Fan

Support: Lite (45)

Surface Layers Top/Bottom: 5/5

Infill: Light

No side doors

Room temp 18 - 21 degrees C

Very light layer of disappearing glue stick

Brand new, just opened Filament

OK, so all you anti-Z-Temp heads look away now....

Check out Part 2 (blue one) - this was printed at the exact same time, in the exact same room as Part 1. My settings for this are as below:

Verbatim PLA: Blue

0.19mm

0% Fan

Support: Lite (45)

Surface Layers Top/Bottom: 5/5

Infill: Light

Side doors On.

Room temp 18 - 21 degrees C

Very light layer of disappearing glue stick

Filament: 1 week old

Conclusions:

For Z-ABS I'm really starting to think that the age/condition of the filament is the key to solving warping issues. This is the second time I opened a new reel of Z-ABS and printed a large part straight away successfully. And then subsequent prints get progressively more warping issues. I know that material storage has been a frequently visited topic on this forum - having said that, can I please bring up this old topic and ask anyone reading this - How to you store your filament? (Room Temp/humidity conditions etc).

I'd also conclude that the room temp NEEDS to be constant. Whether that's constant 20, constant 25, constant 30, just make it CONSISTANT . It helps!  :D

 

OK, onto the PLA - so this doesn't add any weight to the above conclusion about the age/condition of filaments as it's a week old reel that's used about 25% so far - however I just had to post this as the part came out pretty perfect! :lol:.

So there you have it - I've printed a large part without any warping issues :)

Can you guys do the same?!! ;)  Thanks for reading! 

#13

Forgot the pics!

#14

Surprised you've not had good results at 290µ - I use it for everything except parts needing very high detail or really smooth slopes.

#15

Ive notice the color seems to matter, at least for support removal. It may be something you want to track, not sure if it effects warping or not though (note the Inventure supposedly tracks colors, not just material type....)

You might want to also track whether a machine has been fully warmed up before the print, or if has been recently turned on. There will probably be greater intensity air currents, with greater variations in temperature, until the entire machine has warmed up (not just the build plate and hotend).

If you want to get serious about it, I would suggest selecting a representative test print (that's fast  to print) and doing a design of experiments. Right now you have 10 variables, if you assume each variable has two states (which is not true, many have more than that)..... then you have 1024 experiments to do.....  If you do a fractional factorial analysis, you can drastically shrink the number of experiments needed in most cases (you maybe able to reduce it down to as little as 16 experiments, instead of 1024). This is especially useful if you don't care about trying to determine higher ordered interactions between variables, just the main interactions and maybe lower ordered stuff. Zortrax is probably doing a similar statistical approach internally. When you read about doe and fractional analysis, it will look complicated...but it's not really..there are programs to set it up the experiments and calculate the results for you.

#16

Ive notice the color seems to matter, at least for support removal. It may be something you want to track, not sure if it effects warping or not though (note the Inventure supposedly tracks colors, not just material type....)

You might want to also track whether a machine has been fully warmed up before the print, or if has been recently turned on. There will probably be greater intensity air currents, with greater variations in temperature, until the entire machine has warmed up (not just the build plate and hotend).

If you want to get serious about it, I would suggest selecting a representative test print (that's fast  to print) and doing a design of experiments. Right now you have 10 variables, if you assume each variable has two states (which is not true, many have more than that)..... then you have 1024 experiments to do.....  If you do a fractional factorial analysis, you can drastically shrink the number of experiments needed in most cases (you maybe able to reduce it down to as little as 16 experiments, instead of 1024). This is especially useful if you don't care about trying to determine higher ordered interactions between variables, just the main interactions and maybe lower ordered stuff. Zortrax is probably doing a similar statistical approach internally. When you read about doe and fractional analysis, it will look complicated...but it's not really..there are programs to set it up the experiments and calculate the results for you.

Haha I like the way you're thinking! Unfortunately I don't think a representative test part would be fast to print - the part would need to be big and fairly thick (15mm or higher imo) to really challaenge the warping issues with the M200. This is something I wouldn't have the time to do myself - hence forums I guess. But yes, hopefully Zortrax have taken/are taking a statistical approach to figuring out issues like this.

I'm aware of the colour variable as well. I'm presuming it's the emissivity of the material which causes it to react differently to temperature differentials. (Different colours lead to different light reflections, hence different emissivity values.)

The "machine warming up" issue - I'm not sure which parts of the machine you're talking about here. The extruder and bed take minutes to reach equilibrium. With a 9+ hour print I'd say this is negligible no? Again I think it's more about keeping the room at a constant temperature. Any drafts/temp drops can have big effects on the end results (stringing/warping).

#17

Sound good.

For the machine warming up, I was thinking more about the metal casing and internal components, and the psu + steppers heating up. Those probably take longer to warm up than the bed/hot end, until they do I'd expect a change in air currents in the machine due to convection (basically drafts). Even for 9 hour prints, it might still matter as Ive noticed warping happens pretty often early in the print. Just a thought

#18

Marcin works for Zortrax, I think, so his not wanting to discuss Z-Temp is understandable.

So far as storage goes, I keep my filaments in gasketed storage containers made by Sterilite.  Target in the US is one source.

#19

I just read an awesome idea from back in 2010 on an old board. Print a wall/shield around the outline of the part, 1-2 layers thick. Protects the part from drafts, insulates the part slower more uniform cooling, and holds the heat in from the bed (effectively partial heated chamber). I'm going to give it a shot.

#20

What if your anti-draft wall splits?  :P