Ultimaker 2 print strength

[jjl71194 0]

Hi guys,

I've been printing a model that needs to withstand as much force as possible. However, it seems the Ultimaker 2 puts out weak parts in general. I printed out the same model with a Replicator 2X and with the ultimaker and the model the 2X put out is way stronger at 2 shell 15% infill than the ultimaker's at 2 shell 100% infill. Why does the ultimaker's model which is a completely solid piece piece of plastic so much weaker than the 2X's model? I use ABS for all my prints. Is it possible that the Replicator filament is just a stronger abs than the Ultimaker filament or is there another issue here? I got 2.85mm black ABS from 3dUniverse for the ultimaker and 1.75mm makerbot black ABS for the 2X. Thanks

[dodgrr 27]

weird apples to oranges comparison here, you'd need to print both solid or both hollow to get close to an answer, but I don't think that would be an accurate comparison either as one has a nozzle that's roughly %60 smaller than the other, finer woven filament may very well get you more structural rigidity. For the most part, more surface area = more strength.

[danilius 118]

There is no way of answering this question logically. You used two different filaments, so without testing both filaments in both printers, you have no way of knowing which filament is better.

What temperatures did you print at? How fast did you print? What layer thickness? Did you enclose the Ultimaker to keep the heat in?

Basically, you have left out the most important info.

As it is, this sounds a bit like a troll. Having used PLA, ABS and Nylon on the UM2 for functional parts I can tell you that even PLA is a really tough material, and my son slashed his thumb whilst "destruction testing" some PLA which should tell you something as well. What, I have no idea.

[jjl71194 0]

I printed on the makerbot with 110 C bed and 230 C extruder and on the ultimaker I used 105 C bed and 240 C extruder unenclosed. Both were printed at .3 mm thickness and about 60 to 70 mm/s speed. As a rule of thumb I use higher infill percent for a stronger part so I assumed that a part filled completely would have take more to break than a part only 15% filled in. I was just confused why the 15% infill part the makerbot put out was harder to break than the 100% infill ultimaker part.

[danilius 118]

A couple of points. I always print ABS with the bed at 110C and the hotend at 230C for speeds around 30mm/sec and 260C for anything faster than 40mm/sec, and I enclose my printer with A4 paper held together with blutack.

Secondly, there are numerous issues apart from too low a temperature than can cause one printed part to be weaker than the other from the same machine. So, what you need to do if you are really interested in pursuing this is to print two identical parts that are designed to be tested for breaking point.

Mark down ALL the settings, and there are a lot for the Ultimaker. Speed, acceleration and jerk settings, temps, layer thickness and so on.

Print several samples of each from both machines, log the results of testing their break points and then let the community have a look at the data.

This would be useful for everyone.

[gr5 2,167]

I find 240C to be too cold for ABS. The layer adhesion is lousy at 220C but much better at 240C. But still not good enough. If you pay careful attention, it should be breaking along layer boundaries. If that is true then the problem is most likely you should be printing hotter. Not all temp sensors are equal so 230C on mbot might be the same temp as 250C on U2. Also the temp sensors on the U2 sometimes (like 1 in 10 temp sensors) read as much as 10C high (meaning they print 10C cool). So try printing at 250C. Also .3mm is quite thick for a .4mm nozzle. You should consider getting a larger nozzle if you need to print this fast or consider slowing it down a bit. I'm not sure how viscous ABS is at 230C but it seems to me it might be underextruding quite a bit.

Don't do 100% infill. It doesn't really add much strength. Strength of most parts is in the walls, not the center. That's why bones (among other structural things) are hollow. That's why it's okay for a plumber to drill a hole right through the center of a joist but if they cut a notch in the edge the building inspector will make the carpenter put in a new joist.

Of course this is just a general rule - it depends on the shape of the part. For something like a pencil shape 100% infill *is* stronger.

Anyway I suspect you aren't getting the layer adhesion on the U2 that you are getting on the mbot due to insufficient heat.

[gr5 2,167]

Also consider testing the raw filament. I know they are different diameters. The 3mm should be as strong as 2.65 strands of 1.75mm filament - so stronger than 2 strands of 1.75 and weaker than 3.

Test that with a pull test. You could loop the filament over a ceiling hook, stand on a scale and slowly pull harder and harder until it breaks. Repeat with the ultimaker filament. Keep careful records. This will test to see if (unprinted) the filaments are not roughly the same strength. Usually a difference in strength of 2X is considered minor or small. I know - seems crazy. But I've seen some filaments 10X stronger/weaker than others.

[danilius 118]

@gr5 - that's seriously crazy. I know nylon is waaaay stronger than ABS, but to find two PLAs or two ABSs that are wildly different in strength?

I have an ABS filament that at 30mm/sec and at 260C gets badly discolored and the surface finish deteriorates. That's why at 30mm/sec I reduce the temp to 230C.

[gr5 2,167]

I have to admit my ABS prints also tend to be weak. I am an expert at PLA and can get pretty much perfect prints every time but ABS frustrates me and tends to clog my nozzle if I print too hot and tends to have bad layer bonding if I print too cold. I guess I just need to print hotter and faster maybe. When I print ABS I feel like I just got the printer and am helpless. When I say "weak" they are strong enough for most things but tend to be not as strong as my PLA parts. Almost as strong, but not quite as strong.

[JohnFox 25]

No wonder the object is weak, at 0.3mm layer thickness on the UM2 with a 0.4mm nozzle, the layers will hardly bond. It's just like a raft, peals apart easily. For any strength at all you must get each layer to weld to the one below. It won't matter what infill you choose, if you don't get the layers to bond in the first place it will be weak.

Try 0.1mm on the UM2 or use a larger nozzle.

Also 240C may not be hot enough for ABS on the UM2. The machine default temperature is 260C.

 

 

I printed on the makerbot with 110 C bed and 230 C extruder and on the ultimaker I used 105 C bed and 240 C extruder unenclosed. Both were printed at .3 mm thickness and about 60 to 70 mm/s speed. As a rule of thumb I use higher infill percent for a stronger part so I assumed that a part filled completely would have take more to break than a part only 15% filled in. I was just confused why the 15% infill part the makerbot put out was harder to break than the 100% infill ultimaker part.

 

[gr5 2,167]

I would think that a thicker layer would bond better because there is more heat stored in that thicker trace going down on top of the layer below. You don't have to melt the entire layer below - just the top surface.

I also found that printing at 255C was hot enough to cause a nozzle clog and if I print that hot I need to speed up my printing a bit. It was more than a nozzle clog - it was the bottom 10mm of filament that turned into a cooked glob that would not melt.

[JohnFox 25]

The amount of heat is going to be the same assuming constant extrusion but the area of contact between the upper and lower layers is going to be minimal at large layer height. The top layer is going to be a series of cylinders rather than 'squished' on the lower, the only points of contact being the top and bottom of the cylinder. The top layer of a raft works in a similar way so it is easy to peel off.

How do I get images a bit smaller?

Does anyone know the nozzle diameter of the Replicator 2x that jjl is using?

 

I would think that a thicker layer would bond better because there is more heat stored in that thicker trace going down on top of the layer below.  You don't have to melt the entire layer below - just the top surface.

I also found that printing at 255C was hot enough to cause a nozzle clog and if I print that hot I need to speed up my printing a bit.  It was more than a nozzle clog - it was the bottom 10mm of filament that turned into a cooked glob that would not melt.

 

[gr5 2,167]

I see your point - your image shows a layer height equal to the nozzle width. That's why Cura doesn't recommend layer height > 3/4 of the nozzle width (.4mm nozzle means .3mm layer height max).

But if the layer height is too small you have a problem also - so .05mm layer height isn't going to carry enough heat to melt the layer below. I would think a .2mm layer height would give you the most melting without worrying about air/contact. The bead of plastic being laid down would be pretty close to rectangular.

[danilius 118]

I think that there is an issue anyway with the temperature not varying automatically according to speed. This makes little sense, since printing speed can vary considerably during a print (think about printing a cone for instance, where the bottom could be printed at 60mm/s and the top perhaps at only 10mm/s for a quality finish). In this scenario the ideal temperature might be 210C for the bottom and 190C or even less for the top.

I know there are custom firmwares that can do this, but currently I do not have the time to investigate them properly and I do not want to trash my baby by doing something silly.

[JohnFox 25]

Just done a test print at 0.3mm - it's quick!

Quality and strength as expected - poor, starting to see gaps between the layers. Not possible with ABS, it would warp and delaminate. Worth doing with PLA if speed is the only issue.

Certainly explains the original posting about lack of strength.

 

I see your point - your image shows a layer height equal to the nozzle width.  That's why Cura doesn't recommend layer height > 3/4 of the nozzle width (.4mm nozzle means .3mm layer height max).

But if the layer height is too small you have a problem also - so .05mm layer height isn't going to carry enough heat to melt the layer below.  I would think a .2mm layer height would give you the most melting without worrying about air/contact.  The bead of plastic being laid down would be pretty close to rectangular.

 

[KevinMakes 22]

I'm a but confused about the type of strength... There is a few different types so to got a good Idea. Do you wish to: 1.) resist bending. 2.) Bend but retain form. 3.) Have a high surface hardness.? I've found that if you use an infill of about 92% and a layer height that is approx 1/3 of the width with a high head heat, build plate temp of 70C ish for the first few layers, and no more than a 5 second delay between layers (ideally) of the material that it makes a build that retains form after impact. Which is more springy but more resilient. Hope this helps.

[gr5 2,167]

PLA isn't a problem but with ABS it's very common for your printed object to fall apart very easily along layer lines because layers are barely bonding with the layer below. Even with parts that seem okay at first when you stress them they tend to break along printed layers. With PLA this doesn't happen - parts break where the stress is highest. This is because the glass temp for PLA is much lower than ABS so you don't have to heat it as much to get a good connection.

[KevinMakes 22]

PLA isn't a problem but with ABS it's very common for your printed object to fall apart very easily along layer lines because layers are barely bonding with the layer below.  Even with parts that seem okay at first when you stress them they tend to break along printed layers.  With PLA this doesn't happen - parts break where the stress is highest.  

I have found this to be true. Pla is fairly resilient but I squash the layers by using thee overlap feature anyway and I use a very small layer Height so it slightly reheats the previous layer as it lays down the next in succession. however it does make a kindof messy print unless you use like 80-90 infill.