Poor Top Surface

[daveholbrook 0]

 

Hi,

I'm trying to print the best quality test print I can (for my dissertation) on my Ultimaker 2. Below shows images of current best effort along with the settings im using. Any suggestions of how to make this print better would be much appreciated.

Main issues I can see are: bubbling on the top face, and the bottom has a noticeable ridge on it and then dips in before coming back square along the sides.

 

Thanks

 

[IRobertI 520]

The bubbles on the surface is called pillowing (at least I think that's what we've settled on hehe). It's mainly caused by a lack of cooling. What are your settings for the fans? Increasing the infill percentage may also help as it will give the top layer more support.

That the very first layer is sticking out a bit more is called the elephant foot and is caused by the first layer being squished a bit so that it becomes fatter than it should be. I've also noticed that a bed that is too hot makes it worse. I usually put a chamfer on the bottom surface to combat this. 0.5-1mm depending on the print is what I normally use. If you're running the default 70C for the bed I recommend lowering that temperature to 50-60C, I use 60C mostly.

edit: Also, I'd probably print that at 0.1mm instead as you don't really gain anything from the 0.06mm layers with such a basic shape. That should also help with the top surface as it will be easier for the filament to bridge over the infill.

 

[gr5 2,170]

The elephant foot is difficult and I'm not sure what causes it - possibly temperature issues, possibly other factors. As Robert suggests, the most common solution is to compensate in the CAD part. Also different people get different sized "feet" depending on many factors (layer height, temp, speed, fan) so it's hard to fix in Cura. Easier to fix in CAD.

The pillowing as Robert suggests is almost always caused by inadequate fan. Are both of your side fans working? Many Ultimakers came with disconnected side fans. Click "PRINT", choose and model, then quickly go into the TUNE menu and set the fan speed to 100%. As long as you stay in the tune menu the printer will not start printing. If the fans aren't working, try sliding up the black plastic mesh at the top of the head and checking the 3 fan connectors. The wires leading to the 3 fans are red and black.

As robert already said, pillowing can be reduced by thicker layers and also mitigated by having more top layers. But you already have 10 layers! I usually only do 4 and rarely see pillowing. I often don't even do infill (which helps support) and still don't usually see pillowing. So - maybe the fans are off? Also check that they are blowing and not sucking (that was a problem on a UM1 once but I don't think the UM2 fans are capable of spinning backwards).

 

[foehnsturm 969]

My understanding is that the elephant foot is caused by a considerable temperature difference within a small area. While the bottom layers stay at heat bed temperature (and are constantly heated), the layers above are heavily cooled by the fans and therefore shrink. So, if the heat bed temp is too low, the part will lift of. If it is high enough the part will stick but has to give in within the zone near glass transition temp.

This is why I don't have any issues any more since my "cooling" air is almost at the same temperature as the heat bed. Perhaps the main secret of a heated chamber.

 

[gr5 2,170]

Are you saying you don't get the "elephants foot" on your prints anymore?

 

[foehnsturm 969]

Yes, that's why I came to that conclusion.

Having the whole piece at nearly the same temperature during the print really helps.

No heated bed: uniform temp but considerable overall shrinking, low adhesion and high lift-off forces to fight with

Heated bed: good adhesion but high temperature gradient within the part and partial shrinking causing the elephant foot

Heated chamber: good adhesion and uniform temp, very low lift-off forces due to shrinking until the part cools down

[aaronalai 2]

I've been thinking about this a lot, and have seen posts about others trying to implement it in commercial versions, but from what I understand Stratus has the patent on it or something to that nature. I don't have a physical version of the UM2 yet (still in Ready status), but I've been looking at a lot of photos of the machine at as high resolution as I can find, and I think you could get away with sealing up the front with plastic sheet, putting a pyramid trash bag shaped thing on the top to allow the head to move, and some nichrome wire to generate the heat.

Then you run into the problem of how do you keep the crucial temperature differential in the printing head between the melting volume and the sold volume of the filament; I think you can replace the heat sink with a mineral oil heat sink system, and and off-board pump and cooling system for the oil; you would have greater control of the temperature over which the area the heat sink makes contact with the print head structure, because you could alter the temperature of the mineral oil before it enters the heat sink structure. The added mass of the oil and more solid heat sink structure, could be offset by the removal of the 3rd center fan. I am very excited to get my UM2 and have been daydreaming about ideas for it and for prints, even if the aforementioned idea seems convoluted and unnecessary I'm still going to see how I can make this printer as accurate as possible without changing it's fundamental operating system.

 

[aaronalai 2]

Yes, that's why I came to that conclusion.

Having the whole piece at nearly the same temperature during the print really helps.

No heated bed: uniform temp but considerable overall shrinking, low adhesion and high lift-off forces to fight with

Heated bed: good adhesion but high temperature gradient within the part and partial shrinking causing the elephant foot

Heated chamber: good adhesion and uniform temp, very low lift-off forces due to shrinking until the part cools down

 

Does the nozzle of your heated build chamber clog more frequently, or is the idea of a mineral oil cooling system not necessary?

 

[gr5 2,170]

First of all - the heated bed is plenty of heat for a heated chamber. You don't need an additional heater. You need to cover the side and put a box over the top.

The shrinking issues all start around the glass temp so you only have to have the air at around 50C. This should be cool enough to stop jams I would hope. Even if the air is at 40C it is a huge improvement over air at 20C.

Also if the chamber is around 40C and the bed is at 60C then the part is somewhere in between those temps.

I think a more important modification is to keep the servos from getting too hot. On the UM1 this is pretty easy as the motors can be mounted either inside or outside the box with no modifications or hardware changes. They just go either way.

 

[aaronalai 2]

Thanks for the suggestion, after some internet searching it looks like 50-60C is about as high in temp as a stepper should go, and that might be pushing it. I like a good challenge and am interested to see what I come up with. The reason I was thinking of heating the chamber independent of the heating bed was to get the part to be a uniform temp, but I suppose a temp difference of 20C at 40C ambient chamber temp is more effective than a difference of 40C at 20C ambient chamber temp, and probably helps enough.

I was wondering if there were any reports of heated build chambers clogging nozzles more frequently as a function of keeping a lot of filament much hotter than ambient temperature before entering the heated portion of the hot end?

 

[foehnsturm 969]

My steppers are outside (direct drive) and the upper half of the print head and the bowden tube always receive plenty of fresh, cool air. But it requires http://umforum.ultimaker.com/index.php?/topic/3890-the-crossflow-fan-approach/?p=30643. I think there is no patent on the air curtain principle.

 

[daveholbrook 0]

Thanks Robert and gr5. my fans seem to be on when the print is going but if I go into tune I can then turn them upto 100% which has a noticeable increase. Should I leave them at 100% for the entire print? If not then is there anyway of setting them to come on when I require them at certain layers?

Thanks,

Dave

 

[daveholbrook 0]

Also, the elephants foot problem has completely gone by setting the build plate temp to 50 - 60.

 

[IRobertI 520]

Thanks for confirming the elephant foot issue. I can't understand why the machines ship with such high temperature on the bed by default.

As for the fan setting in the tune menu. That change is only temporary and only effects the current print. You can change how the fans behave in cura under the expert settings. By default they are set to come on 100% at 5mm (I think). What that means in practice is that cura will gradually turn them on layer by layer so you may not see the fans go from Off to On sharply but they will come on little by little until they are going full blast.

Personally I turn my fans on a bit earlier than 5mm and then leave them on full blast for the duration of the print. Cooling is more important for smaller objects. On large objects you may get away with no fan at all since each layer takes so long that it is solid by the time the next layer is printed.

 

[gr5 2,170]

Robert speaks truth but more details:

>That change is only temporary and only effects the current print.

It's worse - it might only affect the current layer if you are below the 5mm (or whatever it's set to) threshold because any fan commands in the gcode will override your manual setting until you set it again. And each layer has a slightly different fan speed until you get to the "full" speed (whatever it is set to in Cura).

I believe the whole point for this feature is that for the UM2 the fans were bouncing air off the heated bed and cooling the nozzle too much resulting in underextrusion on the lower levels. I could be wrong about this.

What is the fan good for anyway? With very large prints that take a minute per layer you probably have zero need for it.

For small things where a layer takes only 10 seconds or less it is important to cool the layer below before you add the next layer.

It also seems to help bridging and unsupported overhangs as you want the filament to cool almost instantly as you are laying it down sideways in thin air.

It helps immensely fight "pillowing" which happens on the top layer after covering infill on layers below. Similar to the bridging issue.