Jump to content

jones4642

Member
  • Posts

    15
  • Joined

  • Last visited

Everything posted by jones4642

  1. I've been using Cura 4.13.1 for a while now, and have had great results. However, I'm currently trying to print a rather large part with PET-G and a few modifications to the profile, and I'm getting some odd behavior that is resulting in several additional hours of print time. I've uploaded the .ufp file. I happened to walk into the print room while it was on the 90th layer, and saw the head flying back and forth for no reason that made any sense. Looking back at the .ufp file, I can see what it is doing, but not sure why. It is clearly not efficient. If anyone knows of some parameter changes that might prevent this, please let me know. Edit: it seems to be due to my choice of 999 bottom layers rather than 100% fill. This increased the print time by about 4 hours. 845509185_UMS5_CenterShellV4.3dPETG.ufp
  2. Interesting. It is a lofted part , so may have infinitesimal overhang, on the order of numerical resolution. Looks like I need to spend some time getting to know the blocker function. Now that I'm seeing this part and some similar parts actually printed, I'm not sure that my effort to shell the parts out in CAD is really that helpful. I tried to make the shell with the ideal wall thickness for two layers. This part, with nearly vertical walls is coming out OK, but other parts with more angled walls are a bit of a mess. It ends up trying to make walls 2 plus a small fraction thick, and the walls end up being a bit of a mess - lots of texture in some areas. If I use Cura with 2 wall layers and zero fill, I get really good walls, but it's a bit of effort to get support structure inside the part as needed. I can add in low percentage fill, but again, it's challenging to get material where I want it and still keep the part density low enough to be useful.
  3. Thanks for taking the time. I'm having trouble figuring out why this worked. It looks like you are blocking support starting about half way up the piece, and that prevents all the superfluous support up the side walls. Do you know what that support material was trying to accomplish? Is there something about the shape that would make it want to add support? Oddly, I've printed the same shape, but instead of shelling the part and adding interior support manually, I used Cura to do a low density fill, with one wall layer and something like 2% fill, and in that case it does not try to build support up the sidewalls. The exterior of the part is the same either way.
  4. If you have time, please have a look at the attached model, and see if you can find a way to support the small inset at one end, without building exterior support walls up the exterior faces. I've eliminated the support walls by increasing min support area to 200mm^2, but I still get a prime tower up to the top of the part which increases build time by about 9 hours. The prime tower is only needed for the 1st 1mm of the build. After that it is just a waste of resources. UMS5_Wing V4.2b - OBF part 2.3mf
  5. It sounds easy, but I can never get it to block where I want. In this case I would need to start blocking about 15mm off the build plate. I guess I just need to spend some time with it. Still, this seems like a manual workaround for a software bug.
  6. It's sad to say that I have never gotten the hang of support blocking. The areas I would want to block need support material at the bottom end of the part, but the need stops about 15mm off the bed. The bogus surfaces are another 12 or 15cm off the bed. I'm using the 0.25mm nozzle, so I'm limited to 0.1mm layer height, so that's 1200 to 1500 trips to the prime tower with multiple laps around the tower each time. Here's an example part to illustrate the problem. I have min support size set to 100 which keeps support material out of some small pockets near the base where the support material usually breaks up and makes a bit of a mess out of the part. The real problem is the roughly horizontal hole up near the top of the part. Cura wants to support the ends of the hole, so some minimum value of "minimum support area" is needed, but the prime tower grows up that level whether or not there is support material there. It seems that some of the parameters that affect where support goes will correctly adjust the prime tower height, but other parameters do not. It seems like it should be fixable. UMS5_OBF block 3 V4.1c 1-layer.3mf
  7. Sorry - pinging this one again. I really need a fix for this. It's costing me many days of wasted print time. The logic for determining the top of the prime tower seems to be a bit dodgy. If I can prevent a surface from getting support material using the angle limiter, then the prime tower ends when the support ends, but if I use the "minimum support area" setting, then the prime tower goes all the way up to the surface that was eliminated by the use of that setting, even though there is no support material. Additionally, on the particular part I'm dealing with at the moment, the face that it is trying to support should not even be supported I think. It a hole in the side of a vertical face. I'm only using support from the build plate, and there is no surface hanging over the build plate. It's one of those spots where Cura likes to build a tiny but tall support tower all the way up to the face, but then doesn't actually do anything to support the face. A related issue - every time I get in a situation where Cura is trying to build a tiny support tower, where that tower is the only remaining support, the feeder ends up grinding away the material, I guess due to rapidly shoving it in and out without actually feeding any significant material, and then the printer pauses claiming that it is out of material. I have to unload, cut off most of a meter of material and reload, and then restart, only to have it repeat this 20 or 30 minutes later. Using the "minimum support area" setting, I can generally prevent it from building support structures like this (as long as I actually catch them while I'm in Cura), but I still get the prime tower going to the original height of the useless support structure that I just eliminated. A fix in Cura would be ideal, but if anyone has any suggestions to help me defeat this, I would really appreciate the help.
  8. I think the usual approach is to do this in CAD. Are you trying to get CURA to do this?
  9. This error has persisted through every version of Cura that I have used, up through 4.5.0 now, and it would be great to find a solution for it. The problem comes about when I use breakaway support material touching the build plate only, and I use the "minimum area support" parameter to keep it from building useless support towers up to microscopic facets. The minimum support area parameter does a great job of eliminating these support towers, the problem is, the prime tower continues to build, even though it is just the build material, no support material, above the last level where support material was actually used. While there is measurable material going into the tower, the real irritation is the time involved. On some of my taller builds, this can easily be half a day or more of print time! I really need to find a way to get around this.
  10. Not sure if this is the same problem, but it sounds like a fairly common problem on the S5. Many of us have had issues where the support material oozes during the active leveling process. The problem is that if it is extruder 2 then it is mostly hidden from view during the long leveling process for extruder 1. It just takes a small bump of material to completely thrown off the leveling. If this happens, the nozzle will be too far from the plate and the support material on the first level will not stick. Interestingly, I have two S5s, and only one of them has really suffered from this. Every time I would start a job, I would end up aborting 3 or 4 times until I finally got a good level (hint - if you see the problem during leveling and abort before the leveling finishes, it is pretty fast. If you wait until leveling is finished, abort takes several minutes). This printer had older FW. I just updated FW, and at least on the first print, it seemed to be fixed. Will see if this was a coincidence.
  11. In all versions of Cura that I have used, through 4.4, there is an annoying issue with the support tower. If I use the "Minimum Support Area" setting to keep from using support material in unnecessary places, Cura is good enough to skip the support material, however, it continues to build the support tower up to the full height of the spot where support was skipped. I have some parts where I need support material for the first dozen layers, but then there is no more support material used, and the tower continues for another 200mm+ of height. A huge waste of time and plastic. The location where it would like to build support is a very tall cylindrical hole 3mm in diameter with a flat end to it. If I add a chamfer to the end, and then set the "Support Overhang Angle" to eliminate support, then it not only correctly skips the support material, but the support tower also ends appropriately after the first dozen or so layers. On my latest print, this error would cost me 12 hours and 20g of print material.
  12. Thanks for the quick feedback. Does the expansion primarily control wall thickness? From the discussion above it seemed like a method to get the inner and outer dimensions of a part to match desired values better, but not necessarily adjust bulk scaling issues. If I understand it right, if I printed a 1cm square box, and got an outer dimension that was a little too big and an inner dimension that was a little too small, then a slightly negative value of the horizontal expansion could take care of this. However, it's also my impression that if I printed a 200 a 200 cm box, that the actual error (not the percent) error should be about the same as the 1cm box, and the same horizontal expansion adjustment should work. In my case if I print a 1cm part, and measure it to be 0.997 cm, and then print a 295mm part, that measures as 294mm, would horizontal expansion somehow scale up with the part size and fix this? I'm guessing not, and that I really need to multiple x by 1.0033 in Cura to get the right sized part.
  13. This is a pretty old thread, but perhaps people are still watching it. I have a couple of S5s, and I'm printing really big parts on it that need to mate with parts manufactured on other machines (CNC mill, laser cutting, etc). The problem I am having is that I am getting large enough errors in the X-dim to make parts not fit. I have a part that is supposed to be 295mm in length, but it is really closer to 294mm. As a percent error, maybe this doesn't seem horrible, about 0.33%, but I have bolts that need to attach the printed parts to other parts, and on an M3 bolt, a 1mm misalignment is not workable. Is there a way to perform x,y,z scaling in the printer itself to correct for measured scaling errors, or do I need to manually do this in Cura? It seems like it is really a mechanical issue in the printer, as this would appear to be imprecise math on the motor steps to mm travel. I suppose it is possible that it's print shrinkage or something, which would honestly be more frustrating since that would likely be part and material specific.
  14. Thanks for the feedback. I'll check FW versions. One of the S5s is only a few months old, so I would think newer FW, but perhaps it sat on the shelves for a while before we bought it. Looking at the tensioning gages, they seem to be about in the middle of the slot. Should we just crank them up higher? Is there such a think as too high? If so, how do we know when we hit that? With regard to nozzle blockage, what's the test/treatment for that - or perhaps this is answered in the FAQ.
  15. I apologize if this topic has been addressed already, but I wasn't quite sure what to search for, and didn't find anything relevant. I've been running a lot of pretty long print jobs (3 to 5 days) on a U3 and a couple of S5s, and while the majority of the parts print fine, on a significant portion there seems to be a change in feed-rate at random times through the prints, creating under-filled sections that are weak and easily break. The parts I am printing typically use the 0.25 nozzle and are typically very low-fill parts. The first time I noticed the problem was on a large shell structure that had 4-layer walls, and was otherwise completely empty. About 2 days in, the material feed rate seemed to drop off for a couple of hours (middle of the night), leaving about a 1/2" of build height thinner than the rest. Both below and above the thin section, the print was fine. I just finished a 5 day print, and yesterday there was a similar issue with a very thin line of under-filled material. This was even a lighter construction, with 0.48mm walls (2 layers) and modeled internal structure (as opposed to partial fill). It seems like there is a higher probability of seeing these defects on new rolls of material, although the latest 5-day print was on a roll that was about 1/2 used, so pretty light. I mainly print with PLA and CPE, and have experienced this on both. I have re-printed parts that have failed, and had them come out perfect, so it is definitely not an issue with the Cura file. I'm guessing that this is either an indication of defective material, but I only use Ultimaker material, or the feed mechanism. On all of them, the feed tensioner seems to be about mid scale, and we have never adjusted it. The U3 is a couple of years old, one S5 is maybe 9 months old, and the other S5 is just a few months old. On the first S5, after maybe 5 or 6 months I started to see issues where the printer reported that it was out of material, when clearly it was not. After doing some research on this forum it seemed that this was a known issue, and following advice from other users we disabled the sensor. This of course means that we have to plan large prints carefully to make sure we don't run out of material. The newer S5 had this issue on the first print, so we disabled the sensor immediately. The forums suggest that this is frequently caused by using oversize spools of material, or spools not mounted in the stock position, but as mentioned, we only use Ultimaker material, and only on the built in mount. Any advice, suggestions would be much appreciated, as seeing a print fail 4 days in is pretty disappointing, in particular when it is on a $6k machine.
×
×
  • Create New...