Elephant's Foot

Hello everyone. I have been having some issues with the about the first 8 layers of my prints. The first layer looks normal and then it seems like it squishes down the next several layers until about 8 layers up it seems to work itself back to normal. This results in almost a step from the bottom squished layers to the normal layers. It seems to be a severe case of elephant's foot.

I have tried:
1. All different nozzle heights for the first layer.
2. Changing temperatures
3. Slowing down the first layer.

None of these seem to have any effect. My machine is a Core XY with a cantilevered bed. I suspect something with the bed but I can't be sure.

Thanks.

I presume you've carefully set Z = 0.

You could try using a layer change g-code that moves in Z up and then down by the same amount after the layer change. This makes every Z move approach the setpoint from the same direction and fixed (or masked, depending on how you look at it) a similar condition on my Mendel that I diagnosed as being due to hysteresis/backlash.

I'm using:

G91 ; change to relative positioning
G1 Z1 F2000 ; move up 1 mm
G1 Z-1 F2000 ; move down 1 mm
G90 ; change to absolute positioning

Can you post a picture?

If it is what I'm thinking (imagining), I have had the issue where I was trying to print an object and it caused a slight bulge at the bottom tapering up to normal shape after about that same distance.

What I have read, and confirmed, it is an unlevel bed causing it to squish out to one side at the bottom.

I have seen cases where it tapers like you said Tazz but I have never seen anything that is stepped like this. I have a feeling that is back lash in my Z-Stage. But at the same time when I first finished my machine I saw very dramatic backlash(so much so I could see that the Z rod was spinning and the stage was just sitting for a brief moment while the rod caught up with it) and I made an anti-backlash spring-loaded nut to prevent this.

Is the code you suggested, Boring, basically a Z-lift in between layers? I will try that and see what happens.

Thanks.

Those bottom layers are visibly thinner than the others. Hence the same quantity of extruded goo is being pushed outwards.

Not a slicing thing? Same results with all prints?
Mechanical thing then.

Try adding some weight to your z stage. Give it some pre load maybe. Gravity should be your friend and keep the nut sitting on the screw threads.
(Assuming your Z raises and lowers)
Backlash shouldn't be an issue unless you're lifting Z as part of retraction.
-a

Edit : Just thought - It's not the Z stage sticking a bit at that point and motors skipping steps?

Edited 1 time(s). Last edit at 06/02/2015 04:10PM by alan richard.

Alan, I don't believe it is a problem with slicing as it happens on all of my prints. But I bet you are right on the sticking. Now that I think about I did use bushings on the Z axis stage. I think will redesign my stage to fit some LM10UUs. Maybe they will be smoother than the bushings and allow the stage to slide without sticking or jumping. I will keep you updated.

Thanks.

Just thinking aloud really. I'm not sure.

What is the arrangement of your bed? How are the guides and drive set up?

Can you remove the 'drive nut' (if that's what you have) and move the bed by hand? Is there tight spots around the z=0 mark?

Maybe a tweak of the motor current could save you some rebuilding. The Z motor doesn't see much action so you can afford to have the current on the higher side (within reason)

Or you could sacrifice a few mm of build height by lifting the bed a little. If it is (say) a rough spot on the linear guide. The rest of the print layers look quite even.

Still thinking aloud. Hope it helps.

-a

Bottom layers are smashed.

I get this too! Drives me bonkers. The bottom layers are putting down too much plastic, and what happens with me is it hardens and causes the head to drag across it, sometimes skipping if the belts are loose.

The suggested approach is to add a z offset value of about .15 in your slicer settings. I'm gonna give it a try.

Here's my thinking...The Z axis homes from above. Then all the other moves after that, by default, are approached from below. So, if you have any backlash or hysteresis, you'll have layer heights lower than they're supposed to be (with corresponding squeeze-out) until the backlash has reached a stable position. The z lift on layer change sets it up so every layer change is approached from above (and this still holds if you use a z lift on retraction).

Using a z offset to compensate will just reset the z zero and result in lower first-layer adhesion and poorer layer bonding until the backlash stabilizes.