Jump to content

Recommended Posts

Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

Hi all,

I'm trying to slice this part that have two perpendicular thread. So one is horizontal and come perfect and the other come sliced vertically but generate some unexpected lines at the bottom and top .

I tried Superslicer (just for have a confirmation from other slicer) and it do not produce those irregular lines. Shorted and misaligned .

The mesh file is done with Freecad. (I tried the stock export tool and other mesh plugin and despite it look perfect with no error report at any stl checker its final behavior produce those strange lines)

I tried Cura 5.8.1 (that do not produce error in circular lines at infill 100% like cura 5.9.0 instead do).

Tried also Cura 5.9.0 (same identical result as 5.8.1)

I tried also stock profiles (not altered by me) , same result in that last part of the thread. (top and bottom)

Different layer height produce just more refined errors.

What could it be ?

 

edit:

added sliced with OrcaSlicer (that come made properly)

 

 

 

 

lines01.png

lines02.png

lines03.png

lines04.png

lines-orca.png

Edited by giostark
  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    If you could share the Cura project file (.3mf, set it up then go to File > Save Project) that would help since it has the model and all your printer and print settings.

     

    Threads are really hard to print properly, especially depending on the overall size of your part. What material you print them in is also pretty important: if you just print PLA the odds are reasonable it'll snap at some point in the "screwing it in" process (PETG is better, but printing threads is still hard regardless).

     

    Some settings you need to look at for best chances of success:

    • Mesh Fixes > Maximum Resolution: You want this low. Like 0.01mm low or something. The setting exists because with too much detail older printers couldn't keep up with the rate of gcode commands involved. Not really a problem with modern printers.
    • Quality > Layer Height: Threads are essentially printing a lot of fine detail so you want the settings you'd use for printing that. There's no rule for a minimum layer height for each size of nozzle (just that it should never exceed your nozzle diameter) and it also depends on your printer (it's harder to print shorter layers with a Bowden extruder because you don't have as fine control of the flow rate) but 0.12mm is generally the shortest you have a good chance of pulling off with a 0.4mm nozzle (what most printers come with).
    • Walls > Minimum Wall Line Width: There is a guideline for line width to nozzle size. It should be 60-150% of your nozzle diameter. If you're using a 0.4mm nozzle that means you should set the minimum width to 0.24mm.
    • Experimental > Slicing Tolerance: This is less a quality thing and more of a "trying to make sure it's the right size" thing. If you set it to Exclusive then the print will never be bigger than the model file (it always rounds down when deciding if it should add an extra bit), which is what you probably want. On the default of Middle it can things a little bigger if they round up.
    • Like 1
    Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    Yo Slash ,

    Thanks for your quick and accurate response !!

    This make the best result till now with layer at 0.3.

    Unfortunately the first 2 suggestion do not make any difference. And the fourth also do not affect the result but it have to stay inclusive for matching reason.

    This third setting have to be at 0.25 at least . Over this value it make the same bad result.

    Walls > Minimum Wall Line Width: There is a guideline for line width to nozzle size. It should be 60-150% of your nozzle diameter. If you're using a 0.4mm nozzle that means you should set the minimum width to 0.24mm. 

     

    But with layer at 0.1 (and 0.2) the result is still altered.  (last pic "not-ok")

    i do expect a better slicing at low layer height , instead come better at 0.3. And I should print the part at 0.1 for a better quality of the thread.

    I'm using Nylon on a custom delta , 0.6 nozzle. (I would say well tuned).

     

    line-ok-01.png

    line-ok-02.png

    line-not-ok.png

    test-part.3mf

    IMG_20250123_015227.jpg

    IMG_20250123_015248.jpg

    Edited by giostark
  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    The threads in the file you posted looks fine to me:

    image.thumb.png.7dcd840ab061842ce557b360c6fb5ffe.png

    image.thumb.png.8ed3a776d501df243c75068a4eeccbc8.png

    I'd be quite happy with those if I was printing a thread. A smaller nozzle is always going to better at things like this (with the obvious downside that the whole thing takes ages to print). If you're printing Nylon, make sure it's bone dry, and keep it in a dryer while it's printing if you can. Nylon isn't the easiest to begin with but moisture makes things so much worse. Printing threads in general is still a very tricky business. If you need much better than this your best bet is to go the hardware store and find the right bits and some glue. 🧰

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Thanks Slash 🙂

    The horizontal thread is good but the vertical is none. 

    The  picture seems to show a sort of pyramid and not a circle.

    Yep , the filament must be dried . I'm building a dryer .  My plate is made with inox and I use glue , hot chamber is functional. No mechanical problem at printing it. (last print)

    Look at those picture. This circle is not even a circle . (at 0.1 and 0.2 layer height)

    In the vertical picture , in the hight side is more clear. The perimeter should be a circle and instead it rise too much .

    In the orange picture made with Orca the perimeter is well rounded. As expected.

    lines-irregular-01.png

    lines-irregular-02.png

    IMG_20250111_232706.jpg

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    I print threads with a 1.0m noz. you cant get "round" vertical planes when you are printing horizontal layers. You will have layer lines that create mexican pyrimd steps (with bulges) where each step is a layer height tall and a nozzel width wide in a corner. This is why I print al my threads vertically. When I can't, I alter my design or plate arangement to help. In your case I would roated the model 45° to get both threads "of equal quality" which will be better than your horizontal but slightly worse on your vertical. No idea what that will do to your internals. 

     

    Second option... make the horizontal a sparate component, print it verticaly then attach it. I tend to not print preassurized components (this looks like a difusion valve of some sort) so your intenals may not be alterable. I find internal "ugly threads" are much nicer to seal than external. So printing a female hoizontal is perferable to a male. Then thread a male section in to allow hoses to connect. 

     

    Just my experience. good luck. 

    • Like 1
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)
    15 hours ago, giostark said:

    In the orange picture made with Orca the perimeter is well rounded. As expected.

    True, but in the picture from Orca the threads are only one line thick so it won't be smooth gradients between threads. It just goes _|‾|__|‾|__|‾|_

     

    On 1/23/2025 at 9:48 AM, giostark said:

    lines-orca.png

    If you're screwing this into something hard like metal though, it'll likely just shave off bits of plastic on the way in to make it fit, so being perfectly round (which is impossible in 3D printing and to get close you'd need to use the smallest nozzle you can find). You could also printing in something like TPU, which will have a bit of give to it and sort of mould itself to the other thread.

    • Like 1
    Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    I played a little with Cura and I can summarize as follow:

    The ability of Cura to follow the shell is optimized for the 0.4 nozzle.

    The more the nozzle increase the diameter the more the errors occur.

    Here the same identical part just changing layer height and nozzle dimeter.

    More increase the nozzle diameter more the shape become irregular. In a way that it should not do (I suppose).

    Despite the lines become higher, them should follow the shell perimeter but here this do not happen.

    The error remain with changing layer height (but i post it below for not mixing too much image in a post).

    Seems that the better result , despite just very mitigated but not perfectly rounded , is obtained at layer height of 0.3 with 0.4 nozzle diameter.

    I do expect to see the best result at 0.1 layer height . (at 0.05 the behavior is the same)

    1) Layer height 01 nozzle 0.4 

    2) Layer height 01 nozzle 0.5

    3) Layer height 01 nozzle 0.6 

    rounded-shell-0.1-04.png

    rounded-shell-0.1-05.png

    rounded-shell-0.1-06.png

    Edited by giostark
  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Here the next:

    1) Layer height 02 nozzle 0.4

    2) Layer height 02 nozzle 0.5

    3) Layer height 02 nozzle 0.6

     

    rounded-shell-0.2-04.png

    rounded-shell-0.2-05.png

    rounded-shell-0.2-06.png

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Here the next too:

    1) Layer height 03 nozzle 0.4

    2) Layer height 03 nozzle 0.5

    3) Layer height 03 nozzle 0.6

    rounded-shell-0.3-04.png

    rounded-shell-0.3-05.png

    rounded-shell-0.3-06.png

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Now , to me is clear that something in the code of Cura is not optimized. Some one could bring this to the attention of some developer?  (I'm trying to be cooperative , this program is amazing and the community is too) .

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    I take the time to try other stuff from Cura. I remembered that I had  satisfaction using old versions but for the sake of the innovation I tried the new releases. I appreciate the better ability to manage concentric infill but sometime come at cost.

    This is sliced with stock profile at 0.1 layer height and 0.6 nozzle. It is perfect. So definitively there is something off in the new Cura (my above attempts were made with both custom profile than stock ones).

    I tried those Cura versions:

    4.8.0 > error NOT present

    4.13.1 > error present

    5.x.x > error present

    Let me know if I can be useful to fix this stuff with all the attempts by my side you could need.

    cura4.8.0-0.1-06nozzle.png

  • Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    Every once in a while a model like this comes along.

    Cura says there are no errors in the model.

    Formware.co\onLineSTLRepair says the model has no errors, but added 1134 vertexes.

    MS 3D Builder said there were errors and took a long time repairing the model.

    This is the model sliced in PrusaSlicer.  Along with looking like there are a LOT of errors, the tops of the horizontal threads are completely missing.

    image.thumb.png.761aebfe804b2b2a15267ef775ef9b16.png

     

    So I used the PrusaSlicer repair utility and then sliced again.  I don't like it.

    image.thumb.png.6e622b1a3a4ff775fd1ceedb287558ea.png

     

    This is in OrcaSlicer.  This one looks pretty good.  (I changed the colorization to make it easier to see).

    image.thumb.png.581c394a183ea7f3b5dfebedb8e59671.png

     

    Finally there is Bambu Studio.  This looks a lot like the Orca slice which is expected.

    image.thumb.png.216ed0812f1db788968cc31d2bfb0485.png

     

    So we face the age-old question of "Is it the model, or is it the slicer?"

     

    This is the air cleaner adaptor for a motorized bicycle in Cura 5.9.  It has no errors and the threads look to be formed exactly as @giostark described.  It's just as well that this prints best with the threads vertical.

    image.thumb.png.1244ffc9d335ad0b7f52591dcb9eaeb3.png

    Edited by GregValiant
    • Like 1
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Thanks Greg for your contribution. At least is not my imagination.

    I found other stuffs...investigating and comparing other software. This maybe could help to point where the problem lies.

    I'm trying Superslicer (PrusaSlicer fork) (I'm under Arch linux) and I found that if I leave enabled "print thin wall" and 0% of "overlapping external perimeter" it generated this (horrible stuff that are elevated from the part and out the shape):

    superslicer-tinON-0-overlap-wall.thumb.png.1d203e78e0de4ebc2138d2ae3bbd3bba.png

     

    But If I disable "thin wall" and I rise to 80% (as default for the program) "overlapping external perimeter" it come close to perfect (in Cura enable or disable thin wall do not have effect on the errors):

    superslicer-tinOFF-80-overlap-wall.thumb.png.010e4eab8caf6f2739b223009c73c937.png

     

    If some one is interested I can upload the CAD file (I use FreeCAD) . There is a total of 5 mesh plugin in the program . I tried different of them. Coarse and super refined. All bring the same situation.

    Using back Cura 4.8 bring to me the hell of the concentric lines and the irregular wall. Seems I have to use different programs depending on the object.

  • Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    Typically I would make something like that into an assembly and glue that side piece in so it could be printed vertically as a separate detail.  Your design doesn't seem have enough material behind the threaded end to allow that.

     

    Before you give up, try changing the "Wall Transitioning Threshold Angle" to 30° and see what you think.

    image.thumb.png.495a1ac00b4fba482894130f8dcece4c.png

    Edited by GregValiant
    • Like 1
    Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    Hi there.

     

    What an interesting topic this is. I'll know, -printing treads in objects can be very challenging.

    The slicer and the profile for the printer are key points in here and what Greg say are essential in order to print this.

    We've a special version of Cura that handle such kind of object much better, -"the Cura Master".

    (Generated with Cura_Steam Engine mb-master-20201121).

     

    As profiles for my Ultimaker UM2E+ are really well tuned, I've just used standard settings to slice your object, -so here's what it looks like (no support used for better view of details).

     

    Test_Part.thumb.jpg.1495e8ea4f316d995877b2bc82c06c2c.jpg

     

    I've also attached the gcode file as a reference (for you to study if you like).

     

    UM2E_test-part_Cura_Master.gcode

     

    Good Luck

    Thanks

    Torgeir

    Edited by Torgeir
    Text edit.
    • Like 1
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)
    4 hours ago, GregValiant said:

    Typically I would make something like that into an assembly and glue that side piece in so it could be printed vertically as a separate detail.  Your design doesn't seem have enough material behind the threaded end to allow that.

     

    Before you give up, try changing the "Wall Transitioning Threshold Angle" to 30° and see what you think.

    image.thumb.png.495a1ac00b4fba482894130f8dcece4c.png

     

    Same ugly stuff 🤨

    Cura-wall-transitioning-30.thumb.png.2f2d1b9f4c023eda21653a90dd867b7e.png

     

    I'm trying the fist attempt with SuperSlicer. I added 3 modifier . Two to adjust the infill bridge horror design and the third to remove the thin wall from the vertical thread.

    superslicer-test01.thumb.png.2f69aebe72af432ae1b06cedf38ec0cf.png

     

     

  • Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)
    3 hours ago, Torgeir said:

    Hi there.

     

    What an interesting topic this is. I'll know, -printing treads in objects can be very challenging.

    The slicer and the profile for the printer are key points in here and what Greg say are essential in order to print this.

    We've a special version of Cura that handle such kind of object much better, -"the Cura Master".

    (Generated with Cura_Steam Engine mb-master-20201121).

     

    As profiles for my Ultimaker UM2E+ are really well tuned, I've just used standard settings to slice your object, -so here's what it looks like (no support used for better view of details).

     

    Test_Part.thumb.jpg.1495e8ea4f316d995877b2bc82c06c2c.jpg

     

    I've also attached the gcode file as a reference (for you to study if you like).

     

    UM2E_test-part_Cura_Master.gcode 13.67 MB · 1 download

     

    Good Luck

    Thanks

    Torgeir

     

    Edit :

    I found that release 2020.11.21 but under linux do not work the interface and yep , the thread seems come sliced well and also the horizontal lines are not bad :

     

    20201121.thumb.png.cd401c538ac608d22a9e846bb9238c88.png

     

    But the version 20220507 had the working menu but the fancy lines start again and the concentric lines do ugly stuff. I know I can found compromises . Maybe some version in between.

    20220507.thumb.png.6cd229547fe8cd0871d70a9b35a2a05f.png

    Edited by giostark
  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    I found this software:

    https://github.com/smartavionics/Cura/releases

    The version 4.20.25 as the 4 version seems to produce a good thread buuuuuuut.........

    cura4_20_25.thumb.png.d7250dacfe5d8de66e6016144c9f10e6.png

     

    At the same time have far inferior capability to manage concentric lines (was always like this that I could remember)

    In this Cura 5 is far superior. But maybe the version you have is better built. 🤓

     

    Cura4ugly-concentric-lines.thumb.png.326713bfe82d962dfcd6a1a824abbd35.png

    Cura4ugly-concentric-lines02.thumb.png.1e63c4a5dcca8e4a7abfd71c4b913dc9.png

     

    Still I don't know why it always had a strange finish.. as the 4.8 version I used a lot.

     

    Cura4ugly-lines.thumb.png.fbfcc1435d75d899a4f49063645eab65.png

  • Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    The walls around the center hole are concentric to the hole, but the walls around the "threadform" aren't actually round and the center of their geometry is offset from the centerline of the hole.  There is always that gap to fill.

     

    Cura 5.x eliminated the need for the setting "Fill gaps between walls".  Since SmartAvionics Cura is pre-variable line width, the setting still exists.  If you set it to "Nowhere", you can avoid all those annoying little travel moves.  The nozzle should smear plastic around the gaps as the gaps just aren't very large.

     

    When I have threads at the top of a model, I splice the files.  The base of the gcode might be at 0.20 layer height, but the threads will be 0.10 layer height.  That provides a much better threadform.

     

    This is your test model with 0.2 on the bottom and 0.1 at the top.  I used "PauseAtHeight" to provide the point within each file where they get split.  It does require some checking and maybe minor adjustments to the transition gcode.  Splicing gcodes is something I practice because it's a handy skill to have.

    image.thumb.png.aebc011b5555711baf0ee6bae8fbd9a1.png

     

     

     

    • Like 2
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    Hi there.

     

    So I've been testing with the different Cura versions and found "one"-"kind of" old, but remarked surprising..

    This was Cura version 2.31.  I'll think this was made 2015 ish..

     

    Here's a picture of the "object" sliced with Cura 2.31 (monitored/viewed) by using S3d!

     

    Cura_2_31_Test.thumb.jpg.0592beb8672bb436490fdadeedff7f9d.jpg

    Sliced for UM2E+. Here are the gcode file for anyone interested.

    UM2E_CFFFP_test-part_C2_31_B.gcode  Hmm, -a very big file, -but what a quality (from this time).

     

    To print this object using a 0.6 mm nozzle do not work well.. -Not easily.

     

    So, I've thought that this are a job for a 0.4 mm nozzle..

    Used your profile and changed the nozzle to 0.4 mm -finding this might need some adjustment.

    Sure, using Burt's (Smart Avionics) version of Cura (mb-master-20201121).

     

    By some experimenting the object seems to be something that "might" work?

    I've only adjusted parameters in Cura -in the first three group of Menu; "Quality", "Shell" and "Infill".

    However, I've never had any delta 3d printer, but this should not degrade anything.

     

    Here are the picture of your model sliced using 0.4 mm nozzle with "Cura mb-master-20201121", with your printer selected as active printer.

     

    Cura_Master_Test_0_4_mmt.thumb.jpg.fae8a39f4cae8f920c67e909051047e1.jpg

     

    Here's the object located at the "Zero reference cross", in the left corner of my printer but center of your printer.

    Here are the gcode file: CFFFP_test-partC_Master_0_4mm_C.gcode

    Here are the project file for you to check/adjust or..; CFFFP_test-part_0_4mm.3mf

     

    For testing this project, you might dive the model down so the essential top can be test printed without to much filament loss.  "So set Z height to -15".   Here's a picture of this; Test_Part_04_--Dive_A.thumb.jpg.89e72bf2d7646d66605731216ff359f3.jpg

     

    Here are the same "dived" gcode file viewed in S3D:Test_Part_04_--Dive_B.thumb.jpg.2278c98b418ea616178633799aa69391.jpg

     

     

     

    Now some important notes;-if you look at the gcode files in Cura, you might not be impressed!

    So why, -because Cura's gcode viewer are not very good. Sorry to say so, but you can check in any version you want and you'll see that sliced object form version to version, -may look very different in a given viewer version.

     

    ----------------------------------------------------------------------------------------------

     

    This is too bad, -as Cura are the "NO 1 SLICER" -period! -"And no 2 are S3d IMO".  Maybe the management can give the team a change to improve the gcode viewer? As this will benefit Cura and Ultimaker very -very much.

    (Sorry for the rant. But sometimes we need to do a little..)

    ----------------------------------------------------------------------------------------------

     

    By using this nozzle "0.4 mm" the printing time go sky high, close to 9 hour.

     

    Well, that was just my view of your problem.

    Hope you can use some of it.

     

    Good luck

    Thanks

    Torgeir

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

    • Like 2
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    I always appreciate a good rant.

    • Laugh 2
    Link to post
    Share on other sites

    Posted (edited) · Threads not properly sliced (object not in the shape perimeter)

    Thanks Greg and Tor for digging in this . I'm learning several stuffs here.

    I'm also experimenting the "modifiers" tools to found a good compromise. I switched to 0.4 nozzle.

     

    I redesigned the part for optimizing the flow (it is a part of Chinese immersion pump) . Something cheap , but the original part was so stupidly developed that cracked at the first mount , I took the time to make it work .

    I tried to use the blade in reverse mode.

    Now I'm using the stock orientation and  I optimized the exit hole. (was rounded and small)

    I also removed one thread that was  in practical not useful .

     

    I'll keep the thread vertical because I need the internal cycle perfectly flat.

     

    Cheers guys !!!

     

    side01.thumb.png.6f0d99651f4dc1259d9eb5781b169e7d.pngside02.thumb.png.541697d1bad196a6da1fdca6c6d74577.pngside03.thumb.png.d0f0ad962f4c23095a41705bdbae7afc.pngIMG_20250105_214435.thumb.jpg.8011e7712ee581074386e7dc337498c9.jpg

    Edited by giostark
    • Like 1
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)

    That's a fairly complicated and an interesting part.  I have a story about a fairly complicated and interesting part.

     

    My friend had a 40 foot 1936 Chris Craft cabin cruiser.  He asked me to give it a tune-up.  It had a big in-line 6 cylinder Gray Marine engine.  Piece of cake.

    When I finished, I grabbed my toolbox from the motor compartment, and I stepped on the large, robust, cast iron, exhaust back-flow preventer (that had three water connection bosses sticking out at odd angles!), and I climbed out of the bilge.  As my weight went came onto the large robust cast iron piece - through my foot I felt "snap".

    That snap immediately told me "Greg you idiot - you just broke a large, robust, complicated, probably impossible-to-replace, cast iron part of a 50 year old boat".  I told Art and we began to remove the casting.  Art was hanging on to it as I removed the bolts, and it crushed in his hand.  It was completely rotten from the inside.  I felt better when that happened.

     

    There was a Chris Craft dealer on Jefferson Ave outside the marina.

    We walk in and lay that very large, robust, complicated, broken-into-two-pieces casting on the counter and the counterman didn't ask what it was from, he didn't ask how old it was, he just said "Man, have we been waiting for you." and pointed up on the wall behind him.  On that wall, on a shelf all by itself, draped with red bunting, was the exact casting AND it was in solid bronze.

    1 hour and $350 later (probably $3500 today) it was installed in the boat.  They even had gaskets for it.

     

    Back to printing a replacement:

    Remember that the design constraints for the metal casting were different than the design constraints for FDM.  What you need is a replacement part that fulfills the function, rather than an exact duplicate.  If the horizontal threads were actually a separate piece that could be printed vertically and then bonded into a boss on the housing, many problems would go away.

     

    • Like 1
    • Laugh 1
    Link to post
    Share on other sites

    Posted · Threads not properly sliced (object not in the shape perimeter)
    2 hours ago, GregValiant said:

    That's a fairly complicated and an interesting part.  I have a story about a fairly complicated and interesting part.

     

    My friend had a 40 foot 1936 Chris Craft cabin cruiser.  He asked me to give it a tune-up.  It had a big in-line 6 cylinder Gray Marine engine.  Piece of cake.

    When I finished, I grabbed my toolbox from the motor compartment, and I stepped on the large, robust, cast iron, exhaust back-flow preventer (that had three water connection bosses sticking out at odd angles!), and I climbed out of the bilge.  As my weight went came onto the large robust cast iron piece - through my foot I felt "snap".

    That snap immediately told me "Greg you idiot - you just broke a large, robust, complicated, probably impossible-to-replace, cast iron part of a 50 year old boat".  I told Art and we began to remove the casting.  Art was hanging on to it as I removed the bolts, and it crushed in his hand.  It was completely rotten from the inside.  I felt better when that happened.

     

    There was a Chris Craft dealer on Jefferson Ave outside the marina.

    We walk in and lay that very large, robust, complicated, broken-into-two-pieces casting on the counter and the counterman didn't ask what it was from, he didn't ask how old it was, he just said "Man, have we been waiting for you." and pointed up on the wall behind him.  On that wall, on a shelf all by itself, draped with red bunting, was the exact casting AND it was in solid bronze.

    1 hour and $350 later (probably $3500 today) it was installed in the boat.  They even had gaskets for it.

    What a story..  Not to talk about the ending..

     

    2 hours ago, GregValiant said:

    Back to printing a replacement:

    Remember that the design constraints for the metal casting were different than the design constraints for FDM.  What you need is a replacement part that fulfills the function, rather than an exact duplicate.  If the horizontal threads were actually a separate piece that could be printed vertically and then bonded into a boss on the housing, many problems would go away.

     

     

    Good point.

  • Link to post
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now
    • Our picks

      • UltiMaker Cura 5.9 stable released!
        Here comes Cura 5.9 and in this stable release we have lots of material and printer profiles for UltiMaker printers, including the newly released Sketch Sprint. Additionally, scarf seams have been introduced alongside even more print settings and improvements.  Check out the rest of this article to find out the details on all of that and more
          • Like
        • 5 replies
      • Introducing the UltiMaker Factor 4
        We are happy to announce the next evolution in the UltiMaker 3D printer lineup: the UltiMaker Factor 4 industrial-grade 3D printer, designed to take manufacturing to new levels of efficiency and reliability. Factor 4 is an end-to-end 3D printing solution for light industrial applications
          • Heart
          • Thanks
          • Like
        • 7 replies
    ×
    ×
    • Create New...