Composite Steel Gantry Mill - Seeking Feedback - Page 2

Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 55

Thread: Composite Steel Gantry Mill - Seeking Feedback

  1. #21
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Peter,

    These two simulations show that local displacement on the front of the gantry is 100x greater when resting "on air" than it is with a 2lb/cu ft density polyurethane foam core.

    Mesh Control
    • - Mesh control is set at 10%
    • - Adaptive Mesh Refinement set to medium for displacement, so the solver will keep reiterating the mesh until either results are within 10% of the last iteration or if it has iterated 4 times.


    Constraints
    • Each end is fully constrained


    Load
    • Uniformly distributed load of 1000N to front face


    Results
    • "Resting on air": 144 microns
    • With PU core: 1.5 microns


    Attached Thumbnails Attached Thumbnails Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-12-12-08-a   Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-12-12-11-a  


  2. #22
    Member
    Join Date
    Jan 2018
    Location
    Canada
    Posts
    9
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by PaulAmelang View Post
    If so, I happily take on the challenge!
    I think the reason why the community is excited to give feedback here is because you are on a solid path and have the potential to be very successful with your project with a few minor tweaks to your design.

    I can run for you a sample simulation if you provide a STEP file of your design. A download link on a file sharing service like DropBox or Google Drive via Private message would do. This will be much easier than trying to type out how Peteeng and I are suggesting the simulation be run. I have attached a screenshot of an old simulation I ran that shows the fixtures and forces applied to the bearing carriages. Modelling the whole machine at once is not more accurate. Page 26 and onwards of the Bamburg thesis explains how to add together the stiffness values.

    I get what you mean about "not really aerospace level stuff." Your goals are very achievable. It may be useful to analyze the PrintNC to compare to your proposed build. https://www.youtube.com/watch?v=I4hFd18xF8w

    Quote Originally Posted by PaulAmelang View Post
    Uniformly distributed load of 1000N to front face
    There's your problem. The cutting forces will not be a distributed load. Re-read the Hexcel PDF on page 14. https://www.hexcel.com/user_area/content_media/raw/Honeycomb_Sandwich_Design_Technology.pdf it also says "Uniformly distrubuted load"

    A sloppy-but-good-enough method of accounting for local stiffness in your design is to applied the loads on a block that represents a theoretical small area of the linear rail that the bearings will transfer the strain to the frame.

    Composite Steel Gantry Mill - Seeking Feedback-sldworks_uhq5frbt5z-png

    Attached Thumbnails Attached Thumbnails Composite Steel Gantry Mill - Seeking Feedback-sldworks_uhq5frbt5z-png  


  3. #23
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6254
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hi Paul 1) Having the ends fully constrained is over constraining the model but apples to apples it will be OK I think 2) The distributed load is not the loadcase I'm explaining. Its a local load created by the car loading. Run your original model with air or zero stiffness foam to see what happens. 3) What PU properties do you have? I looked up PU and found some 60kg/m3 properties 2lb/cft is 32kg/m3. So if I halve the 62kg properties are your figures about E=7-10MPa and G=3MPa?

    Re PU foam filling - This is a common practice in boatbuilding and building structures. One issue you will have to deal with is post expansion of the foam. PU is used in boatbuilding for reserve buoyancy. Unfortunately the material will continue to expand for some time (months depending on the environmental conditions) after the initial cure and this will balloon the structure. I have had this happen with expanding PU adhesive in laminated structures I have made (at build 120mm thick 6 months later 125mm thick, 10x12mm plywood lamination) and I have seen it in boats I have owned and repaired. The way to fix this is to constrain the face of importance and post cure the PU in-situ. This is not easy as the foam will generate a lot of internal pressure when heated and being an insulator its tough getting heat into the foam.

    I tried to find info on how F360 deals with meshing but couldn't get access to stuff. Is there a help file you can point me at?

    Paul - I am not being negative about your endeavor, I'm simply pointing out known issues with FE, materials and processes... Peter

    - - - Updated - - -

    Hi Paul 1) Having the ends fully constrained is over constraining the model but apples to apples it will be OK I think 2) The distributed load is not the loadcase I'm explaining. Its a local load created by the car loading. Run your original model with air or zero stiffness foam to see what happens. 3) What PU properties do you have? I looked up PU and found some 60kg/m3 properties 2lb/cft is 32kg/m3. So if I halve the 62kg properties are your figures about E=7-10MPa and G=3MPa?

    Re PU foam filling - This is a common practice in boatbuilding and building structures. One issue you will have to deal with is post expansion of the foam. PU is used in boatbuilding for reserve buoyancy. Unfortunately the material will continue to expand for some time (months depending on the environmental conditions) after the initial cure and this will balloon the structure. I have had this happen with expanding PU adhesive in laminated structures I have made (at build 120mm thick 6 months later 125mm thick, 10x12mm plywood lamination) and I have seen it in boats I have owned and repaired. The way to fix this is to constrain the face of importance and post cure the PU in-situ. This is not easy as the foam will generate a lot of internal pressure when heated and being an insulator its tough getting heat into the foam.

    I tried to find info on how F360 deals with meshing but couldn't get access to stuff. Is there a help file you can point me at?

    Paul - I am not being negative about your endeavor, I'm simply pointing out known issues with FE, materials and processes... Peter



  4. #24
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by peteeng View Post
    Hi Paul - You seem to be set on a course so I shall only respond to significant stuff. 1) edge weldiing of a 6mm box such as you have drawn will introduce significant distortion. I have worked as a welder so speak from experience. By significant I mean relative to building a CNC mill. How will you level a surface as you propose, more epoxy? 2) laminating grade epoxies are thixotroped ie they have additives to make them NOT flow, they develop hot spots when curing and crawl and orange peel. Been there done that...Model a 6mm epoxy bed under your rails and see how they go... 3) you mention strength a lot. There is no element in a CNC machine that is under threat of strength failure 4) every part of a machine deflects and every micron adds up to lots of deflection at the tool. Especially parts that are a long way away from the tool due to magnification of the deflection. You have held the base of your columns as infinitely stiff. This over restrains the model. 4) Polystyrene is 80-90% air so in terms of machine design its air...Its great to be innovative but most innovations fail at its first , second and maybe 3rd attempt. This forum is a shark tank... take note of what people say it will save you lots of $$$ and heartache. 5) no, machines do not have to be made from CI but it is a benchmark and machine builders are moving into concrete and mineral casting. I have modelled a lot of sandwich machine bases and they have not worked out in my simulation experience... I would pick concrete before I went down your current path... Peter
    1) The back side of the gantry does not need to be level. Yes, the front will need to be leveled with epoxy (of course, after the foam is poured into the beam).

    2) All laminating epoxies, or just some of them? I called Precision Epoxy today and they said it will probably be around $100-400, and they are going to send me technical data on their leveling epoxy systems. I will model the machine with the epoxy and see how it goes.

    3) You are putting words in my mouth. I'm concerned with rigidity and to a lesser extent damping, not strength.

    4) I'm aware of that. Thanks. I will model the whole machine before ordering parts.

    4) You are incorrect, polystyrene is not the same as air.

    5) I'm glad people are making epoxy granite machine etc.. I'm sure they get great results, but that is a different animal than what I'm doing.

    Quote Originally Posted by Waynekofuco View Post
    You may be right, but I use epoxy quite a bit for the product I build. I'll offer the following. Leveling epoxies are designed to have super low surface tension so as to find the lowest point of any given area. This is the meniscus effect. This does not account for shrinkage. Leveling epoxies shrink. You can mitigate the effect with fillers like calcium carbonate, but this also mitigates the leveling effect. Yet another issue is the exothermic properties epoxy directly related to volume. In areas where there are variations of the volume of epoxy will change the cure and leveling effect.

    Ill ad another option to your list that I think would yield a 'leveling' outcome.

    4) Do not attempt to contain the epoxy at all. Protect the sides and allow any excess drip/run off. Mind you this wont be a clean process. It will force self surface adhesion making it somewhat convex that will offset the convexing nature of the cure.

    Obviously the above requires that the surface being leveled is concave and not convex. Oh, and lets hope there is no twist in the steel, convex or not...
    Thanks for the info. Yes, for those reasons you listed, epoxy does not cure perfectly flat. With careful application, I still believe it can be an effective way to level a surface. Thank you for sharing your experience and ideas. My favorite idea so far is to temporarily extend the surface with a piece of flashing, pour one thing layer of epoxy down, let it dry, pour another thin layer of epoxy down, then saw off the flashing. Using two or more thin layers will create cause less curing distortions than using one thick layer. Also, if I am going to extend the surface with a piece of flashing, it will be necessary to pour multiple layers so that the final layer is not effected by the small dip where the flashing is connected to the table. For the flashing, I can fabricate a thin triangular tube of 20 ga steel to extend the edge a couple inches.

    Quote Originally Posted by peteeng View Post
    Hi Paul - To get back to your original assertion that a semi-circular section is stiffer then a square section. If an RHS is made the same envelope as the semicircle it is 66% stiffer then the semi-circular section in torsion. It is also stiffer in the other directions. The examples are 200mm high and 100mm deep... but actual dimensions do not matter... Peter
    That is assuming that the center of moments is the centroid of the gantry beam, which is not true in my case.



  5. #25
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6254
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hi Paul - I wish you well with your machine. I'll be looking in occasionally. Peter



  6. #26
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by Blockerra View Post
    I think the reason why the community is excited to give feedback here is because you are on a solid path and have the potential to be very successful with your project with a few minor tweaks to your design.

    I can run for you a sample simulation if you provide a STEP file of your design. A download link on a file sharing service like DropBox or Google Drive via Private message would do. This will be much easier than trying to type out how Peteeng and I are suggesting the simulation be run. I have attached a screenshot of an old simulation I ran that shows the fixtures and forces applied to the bearing carriages. Modelling the whole machine at once is not more accurate. Page 26 and onwards of the Bamburg thesis explains how to add together the stiffness values.

    I get what you mean about "not really aerospace level stuff." Your goals are very achievable. It may be useful to analyze the PrintNC to compare to your proposed build. https://www.youtube.com/watch?v=I4hFd18xF8w


    There's your problem. The cutting forces will not be a distributed load. Re-read the Hexcel PDF on page 14. https://www.hexcel.com/user_area/content_media/raw/Honeycomb_Sandwich_Design_Technology.pdf it also says "Uniformly distrubuted load"

    A sloppy-but-good-enough method of accounting for local stiffness in your design is to applied the loads on a block that represents a theoretical small area of the linear rail that the bearings will transfer the strain to the frame.
    Blockerra,

    I thanks for the feedback.

    What makes my original simulation results unrealistic, besides that the Y-beams are overly constrained? It seems like there will be less room for error if I set up the simulations based as closely as possible on how the machine will actually be stressed during operation. I would be more interested in going the direction that Peter is suggesting, adding in the table and workpiece.

    Those last simulations were in response to Peter's comments that the PU foam would not provide any local rigidity to the face of the x-beam. I'm aware that the x-beam won't actually be loaded in that manner.



  7. #27
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1523
    Downloads
    2
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Have not read past first post yet....

    Make your Z travel longer (this does not mean increasing clearance under gantry)

    You want to be able to get the tool tip up to the gantry clearance level.

    Example:
    4 inch thick workpiece/stock
    4 inch long tool needed
    You need 8 inches Z travel to get the tool over the workpiece and to be able to reach the bottom of the workpiece



  8. #28
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6254
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hi Blockerra - To help with the inverse stiffness thing. The inverse of stiffness is compliance. Stiffness (k) and compliance (1/k) if you are familiar with resisters you will know that if you put resistors in parallel the total resistance is the sum of the resistances. If they are in series the total resistance is the invert sum of the resistors. This is called the resistor analogy. See attached for more explanation. As the elements of a normal machine are in series (z axis then saddle then gantry then base then fixture you sum the compliances to figure out its total compliance then you invert to get its stiffness. You can imagine that if you have 4 very stiff springs in series then a soft one, the soft one will dominate the deflection and conversely if you have a highly restrained part/loop it can dominate the result giving an incorrect high stiffness. So to work through this in FE you restrain the bearings at the z axis and get that loops stiffness, then restrain it at the saddle cars and get that loop, then restrain at the gantry cars etc etc. Then you can figure the most compliant loop and that gives you where you need to put the most effort. Hope that helps... Peter

    Attached Thumbnails Attached Thumbnails Composite Steel Gantry Mill - Seeking Feedback-stiffness-jpg  


  9. #29
    Member
    Join Date
    Jan 2018
    Location
    Canada
    Posts
    9
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by PaulAmelang View Post
    What makes my original simulation results unrealistic, besides that the Y-beams are overly constrained?

    Your simulation does not represent how the strain is transferred from the tool to the workpiece. Pete and I are working with a few different assumptions that change the way that the simulation would be run. I'm sorry that it is confusing. I don’t think Pete and I really disagree, we are just going about it a slightly different way. Pete seems to be more knowledgeable than me in the topic.

    Your first simulation treats the linear rails and bearings as solid, they are not. The way to solve this is to break the simulation into parts and apply loads and fixtures to the places where the strain transfer begins and ends.

    Your simulation ignores degrees of freedom. I don’t think anyone has mentioned this yet. The solid rails in your simulation are adding stiffness in the directions that they are free to move. If you disagree, please explain some of these concepts? the tangential stiffness of a radial ball bearing, the longitudinal stiffness of a linear bearing. I'm poking fun, sorry. =)

    The local stiffness problem that has been discussed earlier. It would be easiest to explain if I could just run a simulation with your model.

    If you look at the VMC frame simulation that I posted earlier, the green fixed geometry (constraints In F360?) is applied on the bearings nearer on the strain path to the workpiece, so it is fixed. The force (purple) is applied to the bearings as a two-force pair that is equivalent to a hypothetical force acting upon the cutting tool.

    The way that fixtures/constraints are applied over-constrains the model. There definitely would be cases where the way you did it would be correct. It is my opinion that you will have difficulty achieving a true “apples to apples” comparison if that is how the simulation is run.

    Imagine you put a big spring on a table and rested a mass on top and measured the change in length of the spring. You could use hook’s law to calculate the stiffness of the spring. Now imagine that the table is very flimsy and the mass is heavy and you instead measured the displacement of the top of the spring relative to the floor. In this analogy, your simulation is using the data in the first case as if it applied to the second case.

    Doing the simulation as described above will give you a very high number. For example, one of the components of my design is k=17857 N/um. You then sum these numbers as compliances as Pete explained before (also Bamgurg p26 onward) to get your real number for X, Y and Z stiffness/compliance.

    This real compliance can be compared to your target value so you can decide if a component is actually a weakness or a trade-off. Like was said before, the compliance equation is very unforgiving if a single component is not sufficiently rigid.

    If you would permit me to make a suggestion: Do the compliance math in reverse so you know what value of stiffness/compliance for each component will give you the system stiffness/compliance that you are looking to achieve. Again, Bamburg page 26 has an example of this. In his example, each component needs to be at or above 400 N/um in stiffness to achieve a system stiffness of 50 N/um.

    Summary:

    • Your simulation does not represent how the strain is transferred from the tool to the workpiece
      • Your first simulation treats the linear rails and bearings as solid, they are not.
      • Your simulation ignores degrees of freedom.
      • The way that fixtures/constraints are applied over-constrains the model.




  10. #30
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6254
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hi Block - Degrees of freedom... Yes the bearing has a freedom in the rail direction (called the rolling dirn) so can't actually create a force in that direction. The force/reaction for this part is transferred to the ballscrew or drive system. At the "structural" level of modeling (the autospeller put in meddling and that's probably a better description) you can neglect that issue in the knowledge that its not right. But at some point once you resolve the structure a bit then you put the motion system in and can correctly capture that loadpath. If the car is modelled "soft" say 70GPa vs 200GPa then it will be closer to the actual car stiffness (but it does not capture the slipping/freedom of the car). There are a couple of places that explain how to calibrate a car block so it behaves like a real car (see Bamberg thesis and my threads). I believe in F360 you can allow a surface to slide in which case you need to model the drive system to stop it sliding off the end of the rail and to capture the more compliant loadpath... To summarise 1) cars are bonded to the rail in Round 1 or rough modelling 2) once the structure and objectives are better understood then motion parts can be included if your FE can allow sliding then introduce it. R2 or R3 3) also with rails you have an asymmetric connection in reality. When the rail is pushed into its foundation its as stiff as the foundation when it lifts its as stiff as the bolts holding it which maybe compliant. So FE models are developed in a series of rounds that start simple and get more complex as details are needed or understanding is grasped... That's why I usually say the real article will be 50% the stiffness of the model... once bolts, friction, non linearities, membrane action, hysteresis and all those realities kick in... I have Inventor (but rarely use it, mainly for its form finding function) and if F360 is as good as Inventor FE has, then all of those things can ultimately be captured in the model in Round 76 to 100...Peter

    https://en.wikipedia.org/wiki/Aircra...se%20to%20tail. same names are used for land vehicles and boats. In a boat the roll axis is called surge...



  11. #31
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by catahoula View Post
    By doing it your way, with a very thick core, you're relying on the foam a whole lot more, such that a vertical load on the middle of the table is essentially only resisted by the compressive modulus of the foam.
    I'm sorry, I said you are wrong earlier, but you are correct. Even though the foam has a low compressive modulus, I think it will be okay since there is a large surface area and any load it sees will be well distributed. Simulations should show whether or not this is a problem.



  12. #32
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Peter,


    PU properties for 2lb/cu ft are elastic modulus = 1500 psi and shear modulus = 380 psi. I'm sorry, the PU properties I was using were incorrect. The last simulation I posted, run with the correct PU material data, is 14 microns of displacement.


    In F360, the mesh can be set up different ways. Here is a link to the documentation for the "adaptive mesh refinement" feature: https://help.autodesk.com/view/fusio...NEMENT-CONCEPT
    Basically, when the simulation is being run, the mesh will be reiterated until the results are consistent.

    Thanks for relating your first hand experience with PU. I will be using a rectangular gantry after all which will allow me to use XPS instead. If someone really needed to use PU in a gantry, probably the best way would be to cure it in an oversized mold, put some heat lamps on that baby for a week, then take it out and cut it down to shape.




  13. #33
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Blockerra,

    I understand the concept of local rigidity. I admit that I did not initially. Essentially, there is a limit to how much you can effectively expand the cross section of something while retaining the same amount of mass. A huge pipe made out of thin aluminum foil would not be able to transfer a local load and would deflect/fail locally. I'm willing to bet that using a core of some kind, even if it is a relatively weak material, can significantly expand the limit of how large a cross section you can effectively use.

    Because my gantry is so tall, I believe that it is mostly seeing bending rather than torsional loads. The bottom rail of the gantry is taking most of the load.

    Quote Originally Posted by Blockerra View Post
    Summary:

    • Your simulation does not represent how the strain is transferred from the tool to the workpiece
      • Your first simulation treats the linear rails and bearings as solid, they are not.
      • Your simulation ignores degrees of freedom.
      • The way that fixtures/constraints are applied over-constrains the model.
    No matter how much you adjust the settings of a simulation, it will always be a simplification of real life. The question is not whether or not the settings are perfectly realistic (because they never can be) but rather whether or not they are good enough to generate useful data.

    If you are going to simulate the components separately, how do you capture the exact force vectors? Summing compliance is one thing, but how do you know the exact direction of different forces being transferred between separate components?

    Quote Originally Posted by Blockerra View Post
    Imagine you put a big spring on a table and rested a mass on top and measured the change in length of the spring. You could use hook’s law to calculate the stiffness of the spring. Now imagine that the table is very flimsy and the mass is heavy and you instead measured the displacement of the top of the spring relative to the floor. In this analogy, your simulation is using the data in the first case as if it applied to the second case.
    Kind-of, sort-of. That is an extreme example. The results would be unusable if you constrained something that moved a lot in real life. The results could be usable if you constrained something that only moved a tiny bit in real life.

    I know I am very stubborn, but I am extremely appreciative of everyone's help. I am certainly interested in knowing how my simulations can be run more effectively. I really do appreciate negative feedback, I just need to be able to wrap my mind around it.

    Last edited by PaulAmelang; 05-14-2021 at 09:00 PM.


  14. #34
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Here is an update:

    #1 New Gantry
    I switched to a 12"x18" rectangular tube, still made out of 1/4 for now. I ran some simulations comparing the semi circle gantry to a 2:3 rectangle of the same outside perimeter, and the rectangle faired a little better. Since my gantry is relatively tall, the bottom rail is taking most of the load and it is mostly a bending load. The rectangular gantry has a strong lower end and I think this is the main reason that the rectangular gantry did better in this case.

    #2 Switching to all XPS polystyrene cores instead of PU
    The main reason I chose with PU previously was because of the odd shape of the gantry. Now I can use XPS and it will be easier than having to deal with the PU.

    #2 Added epoxy to the model
    Added a 1/4" layer of epoxy to the table top, Y beam faces, and gantry face. Precision Epoxy says on their website that a 3/8" pour is acceptable too. I'm still waiting to hear back from Precision Epoxy with their data sheets. The guy said on the phone that they don't have data sheets for each epoxy, but he could get me the numbers I need. Meanwhile, I called West System to ask about using their epoxy for making a surface plate, and they referred me to their other company, entropyresins.com. Entropy makes a casting resin with a viscosity of 270 centipoise that might work well. It looks like if getting a specialty leveling epoxy is out of the budget, a casting resin would be the next best thing. They are very low viscosity and have a long cure time, which I assume would mean less heat distortion during the cure.


    Simulation Details

    Constraints
    One vertex on each bottom corner is fuller constrained. I'm sure there was a better way to do this but....

    Load
    1000N into the end of the spindle
    1000N opposite direction into a steel block attached to the table (I know there's got to be a better way to do this)

    Notes:
    Carriage are modeled at 70 GPa modulus instead of 200, as per Peter's suggestion. Epoxy material properties are from Entropy Resins casting epoxy. I repeatedly tried add the ball screw assemblies and making the carriage/rail contacts slider joints instead of bonded, but my computer just couldn't handle it, even though the simulation was run in the cloud.

    Deflection:
    15.17 microns at end of spindle
    4.01 microns in center of "workpiece"

    Takeway
    The first pic is of deflection, the others are of strain. You can see some strain in the epoxy underneath all of the rails. There is a good amount of deflection in the table, I should probably beef up the table skins, but I would be exceeding my weight quota. I've really got to wrap this project up, as it is just a means to the end of getting my business of the ground.

    Attached Thumbnails Attached Thumbnails Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-15-12-58-a   Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-15-1-00-a   Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-15-1-01-a  


  15. #35
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Drilling Simulation


    Settings

    Same as last simulation, except the loads are set up in the Z axis, 1000N up into the spindle and 1000N down into the workpiece.


    Deflection

    13.29 microns in spindle
    52.61 microns in workpiece


    Takeaway

    The table core is compressing and not transferring the load to the bottom skin. The displacement is not acceptable. A couple options come to mind: 1) reduce thickness of table, 2) add supports/ribs in the foam core, 3) use higher density foam. I'll simulate some different options. I'd like to keep the table thick if possible by beefing up the core, but we'll see how the simulations turn out. Using a higher density foam is probably not a good direction to go because of the marginal difference it would make and because only the light density XPS is available to me locally. I wish I had taken some pictures of the stress, but now F360 is running another sim.

    There is a feature in F360 called "remove rigid body modes" that allows simulations to be run without being fully constrained or without being constrained at all, under certain conditions. I will be utilizing that feature when I can from now on.

    Attached Thumbnails Attached Thumbnails Composite Steel Gantry Mill - Seeking Feedback-screen-shot-2021-05-15-11-05-a  


  16. #36
    Member Waynekofuco's Avatar
    Join Date
    Mar 2021
    Posts
    41
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by PaulAmelang View Post
    Here is an update:

    #1 New Gantry
    I switched to a 12"x18" rectangular tube, still made out of 1/4 for now. I ran some simulations comparing the semi circle gantry to a 2:3 rectangle of the same outside perimeter, and the rectangle faired a little better. Since my gantry is relatively tall, the bottom rail is taking most of the load and it is mostly a bending load. The rectangular gantry has a strong lower end and I think this is the main reason that the rectangular gantry did better in this case.
    Hi Paul,
    Im following this thread w great interest because of all the well seasoned feedback. I'm a noob, so bare with me. The gantry tube is still steel? 12x18" recangle tube? Where are you sourcing this? Fyi when I did my steel frame I was amazed at how loose the dimensional tolerances for structural steel are. I fear you may be modelling/simming materials that either unavailable or with tolerances so wildly loose to make it a mute point.

    As a practical/economic matter I purchased my steel (non specific structural 1/4" wall ) from a local supply that was a lot cheaper than any online source I could find. I purcahsed Jan of this year and the sales guy told me I was lucky that I was purchasing now as steel prices are jumping 40% Feb. Im in USA and dont know if the spike is world wide or not, but you may want to try and source/costing in your locality now rather than later.

    Again just my exp as novice machine builder.



  17. #37
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hi Wayne,

    I think I have less experience than you do.

    The gantry and Y-beams will be fabricated out of 1/4" steel plate. The way I have it set up, the Y-beams and gantry can be cut from one 4x8 sheet. If I bought an extruded tube, I would have to buy the entire thing, probably a 20' or 24' section. Also, fabricating the gantry and y-beams from plate allows me to build them around XPS cores. My options for attaching the pieces are either edge welding or bolting steel angle to each corner, or epoxy. I believe edge welding will be acceptable for the gantry since I will connect the pieces before leveling the face. Except.. I'm pretty sure doing so would melt the foam core, so I guess that idea is out. Well, there are multiple ways to skin a cat. I believe the tolerance for 1/4" hot roll steel is about +/- 0.03". I believe Precision Epoxy says their epoxy will level to within 0.003" (I didn't get that figure from their website, I saw it second hand on a forum). I plan to level the table top and the faces of the Y beams and gantry with epoxy. Simulations seem to show that you can still achieve acceptable rigidity with a layer of epoxy.

    Luckily I live in Texas in the country, and the local metal market has outstanding prices, although they only carry hot roll "barn steel", and of course the prices have gone up a lot lately on steel everywhere. If I need aluminum or any special alloy I can get it from Austin or Houston at a price better than is available online. I will compare the prices of aluminum and steel before making the purchases, because I don't think aluminum prices have shot up as much as steel, but I doubt that the disparity would make aluminum the more economical option for the bulk of my materials.


    I plan to make this baby with a circular saw, homemade saw track, drill gun, some clamps, and some measuring tools like a caliper, square, 123 blocks, ruler, and indicator. I've designed it to be able to be assembled precisely this way.

    About the epoxy meniscus: If the restraining lip is above the epoxy line, the meniscus will be concave. If the meniscus is below the epoxy line, the meniscus will be convex. If the lip is equal to the epoxy line, the meniscus should be minimal. My best bet is probably to make a perimeter (some thin sheet metal) that sticks up precisely 3/16 around the perimeter, and then calculate exactly how much epoxy is needed to raise the surface that amount. Of course it won't be perfect, but I bet I could get it pretty close and reduce the meniscus effect considerably this way. In order to minimize the variable of how much epoxy is left in the bucket, I would want to use a spatula to get out every drop I can. It is also standard practice to mix the resin system in a container, then transfer it into a second container to mix it again before pouring it, that way there is less of a concern of any of the resin/hardener not mixing completely.

    Last edited by PaulAmelang; 05-15-2021 at 03:36 PM.


  18. #38
    Member Waynekofuco's Avatar
    Join Date
    Mar 2021
    Posts
    41
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Quote Originally Posted by PaulAmelang View Post
    Hi Wayne,

    I think I have less experience than you do.

    The gantry and Y-beams will be fabricated out of 1/4" steel plate. The way I have it set up, the Y-beams and gantry can be cut from one 4x8 sheet. If I bought an extruded tube, I would have to buy the entire thing, probably a 20' or 24' section. Also, fabricating the gantry and y-beams from plate allows me to build them around XPS cores. My options for attaching the pieces are either edge welding or bolting steel angle to each corner, or epoxy. I believe edge welding will be acceptable for the gantry since I will connect the pieces before leveling the face. Except.. I'm pretty sure doing so would melt the foam core, so I guess that idea is out. Well, there are multiple ways to skin a cat. I believe the tolerance for 1/4" hot roll steel is about +/- 0.03". I believe Precision Epoxy says their epoxy will level to within 0.003" (I didn't get that figure from their website, I saw it second hand on a forum). I plan to level the table top and the faces of the Y beams and gantry with epoxy. Simulations seem to show that you can still achieve acceptable rigidity with a layer of epoxy.

    Luckily I live in Texas in the country, and the local metal market has outstanding prices, although they only carry hot roll "barn steel", and of course the prices have gone up a lot lately on steel everywhere. If I need aluminum or any special alloy I can get it from Austin or Houston at a price better than is available online. I will compare the prices of aluminum and steel before making the purchases, because I don't think aluminum prices have shot up as much as steel, but I doubt that the disparity would make aluminum the more economical option for the bulk of my materials.


    I plan to make this baby with a circular saw, homemade saw track, drill gun, some clamps, and some measuring tools like a caliper, square, 123 blocks, ruler, and indicator. I've designed it to be able to be assembled precisely this way.

    About the epoxy meniscus: If the restraining lip is above the epoxy line, the meniscus will be concave. If the meniscus is below the epoxy line, the meniscus will be convex. If the lip is equal to the epoxy line, the meniscus should be minimal. My best bet is probably to make a perimeter (some thin sheet metal) that sticks up precisely 3/16 around the perimeter, and then calculate exactly how much epoxy is needed to raise the surface that amount. Of course it won't be perfect, but I bet I could get it pretty close and reduce the meniscus effect considerably this way. In order to minimize the variable of how much epoxy is left in the bucket, I would want to use a spatula to get out every drop I can. It is also standard practice to mix the resin system in a container, then transfer it into a second container to mix it again before pouring it, that way there is less of a concern of any of the resin/hardener not mixing completely.
    I started in much the same manner with my last machine. I wound up spending close to $10k in welding, measuring and fixturing equipment before spending any money on the mill components. For me, its money well spent as Ive always wanted to learn these skills using these tools/instruments . My wife may not agree, but I'm in a position to qualify this as business necessity. Wishing you the best, it will be quite and endeavor, hopefully just as satisfying when it done.

    If there was one thing I wish I would have spent more time considering, it would have been enduse/application of the machine. While my creation works exceptionally better than my Amazon/Ebay 6040 router in every respect, it has a major mechanical flaw. Initially I envisioned fixturing everything to a fixture plate much like I did on the router it replaced. Fast forward to putting on a new 6" vice (tegara 660UR, was gonna get the Kurt, but wanted the reverse orientation) and it works perfectly for 99.9% of the designs I want to mill. Unfortunately, it skews the Z work area to the weakest point of the Z axis...

    Regarding the epoxy. Ideally yes, practically you'll want to 'over run'. Other key practical issue;
    - You really want to make at least +20% needed as you want to transfer containers multiple times to avoid isolation of part A or B which creates loss before the pour. If for any reason there is not complete mixture, the pour will be soft and not fully cure. This typically happens when you mix both parts in a single container. Part A is heavy and separates down while the catalyst floats. The only way to ensure no separation is to mix in a container then transfer back and forth in multiple containers before pouring. This isn't such an issue when making epoxy granite using fillers, but when the epoxy itself is the sole substrate of a mechanical assembly, the epoxy mix becomes critical.

    ADDENDUM; And bolt anything you can and don't weld if possible. I spent a lot of money on 1/4" flat stock for the mating plates and even more on 246 block to fixture them flat/square to each other. Everything looked good tacking, but running any sort of a bead distorted things substantially. I spent many months researching weld sequencing and tech. End of the day, all of my weldments required major accommodation to tram things...

    I'm planning another mill and after the steel attempt epoxy/concrete granite is the path Ill pursue....

    Last edited by Waynekofuco; 05-15-2021 at 05:24 PM. Reason: addendum


  19. #39
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6254
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Hello Paul - So many things to do... Do not use "remove rigid body moves" as a general strategy. This is to fix issues with models that can't be fixed any other way. It will introduce uninterpretable results if used generally. Load the tool end and restrain the vice surface. The machine is floating in space as you have not included gravity. A CNC should be so stiff that gravity can be neglected. The Z axis can sag a bit so gravity can be included once you get down the track a bit further. If you include gravity then the feet of the bench need to be constrained but allowed to slip around as if they are on a greasy surface. Peter

    To discuss foam a bit more. 1) The density foams you are discussing are generally considered not to be structural, they are insulation grades 2) I would consider using 300kg.m3 pouring PU or 400kg for your fill. It expands much less and is much stiffer 2) Trying to wedge PS foam into your box and get it to couple will be a mission 3) Mills throw around a lot of coolant. You will be drilling holes into the foam through the plates. Oil will slowly seep into the foam and interfaces and over time will degrade the core. Ask boatbuilders how many times they have fixed cores at penetrations due to water damage. Oil will seep further then water.... so you need to seal the back of the holes somehow if you go down this path. Keep at it... Personally I think you should consider a UHPC machine, I see it as an easier path to control and Wayne and others have been through the steel process have come to that conclusion as well. Industry is coming to that conclusion as well. And concrete is easy to do in your garage and its cheaper then steel.... The tool list you have is fine for making MDF moulds and will produce parts that are much more accurate then welded steel and it will be damp.... Peter

    After lots of research here's the material to use-----

    Attached Files Attached Files
    Last edited by peteeng; 05-15-2021 at 05:48 PM.


  20. #40
    Member PaulAmelang's Avatar
    Join Date
    Apr 2021
    Posts
    32
    Downloads
    0
    Uploads
    0

    Default Re: Composite Steel Gantry Mill - Seeking Feedback

    Wayne,

    Thanks for relating your first hand experience. I am looking at about $2k-$2.5k for everything I need to buy right now, since I already have the spindle, motors, and electronics. I will be sure to account for fixtures before finalizing the design/buying parts. I will be using a vacuum chuck for milling sheet metal, but I will need to think about a vice for working on other items.

    I have a simple and effective idea to precisely measure how much epoxy I'm pouring:
    1) Measure weight of bucket alone
    2) Mix up a epoxy (including some extra) a couple times in separate buckets
    3) When I'm ready to pour, I can measure weight of bucket with epoxy and subtract the bucket weight to find how much epoxy is in the bucket.
    4) Pour most of the epoxy, leaving some in the bucket
    5) Continue to pour tiny amounts of epoxy, weighing the bucket after each time, until the exact amount of epoxy is poured to fill the leveled surface up to the measured line. I will be able to pour the correct amount within a margin of however accurate my scale is.
    6) Assuming that I have precisely installed the lip, the epoxy should be damn near flush with the lip.

    BOOM, no meniscus.



Page 2 of 3 FirstFirst 123 LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Composite Steel Gantry Mill - Seeking Feedback

Composite Steel Gantry Mill - Seeking Feedback