Welded Steel Frame VMC-Interpreting simulation results


Page 1 of 2 12 LastLast
Results 1 to 20 of 40

Thread: Welded Steel Frame VMC-Interpreting simulation results

  1. #1
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Welded Steel Frame VMC-Interpreting simulation results

    Building my new machine, based on waterjet scrap and surplus industrial components. (as you do)

    I modeled the column weldment in solidworks, as the materials to hand drove the construction of the column. I've done a frequency analysis here and am having a tough time interpreting the results.







    The column is welded A36 steel, the base plate/mounting plate is 0.750" thick and 14.5"x12", the braces around the base are all 0.375" thick, the column tubing itself is 10"x6" rectangular hollow section with a 10mm wall thickness. The linear rail mountings are 1" plate steel, T-welded to 0.625" spacers to make room for the ball screw nut.

    The column has a 3/8" thick top cap welded into close up the top end of the tubing, and all welds are V-grooved for full penetration TIG roots and 7014 cap passes. The top cap has not been welded in place yet, and I have not added the mounting pads for the ballscrew or the tool changer.

    I have not started on the damping yet-I am concerned mainly with designing a rigid structure first, and pushing the resonant frequency up as high as I can get it before starting the damping process (which will bring the resonant frequency down as a byproduct of adding mass) so that I end up with a well damped rigid structure with as high a resonant frequency as possible.

    Thanks for any advice-especially on how to set up and run the simulations.

    Similar Threads:


  2. #2
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - Try using the thread tools for images vs links. Links will also expire in a default period. The images you linked to don't open. Interpreting a modal analysis goes like this. The structure wants to vibrate at its "natural" frequencies. This is local and global. Local free edges are the usual culprit, like brackets end of extrusions etc. Then you start getting global modes like the entire gantry wobbling. If there is a driving frequency in the machine, say at feed speed of the process the stepper pulses at a certain freq. This pulse can excite the structure if its the same freq. The tool cutting "bumps" the structure at a certain freq and this causes "chatter". Again if the Z axis has natural freq at this bump rate it will vibrate and make poor surfaces.

    Is this the sort of info you need? or more along the lines of how to do it? For a modal analysis you don't need to apply loads but you do need to restrain it in the intended way as this can affect the response. A machine has to be very stiff. If its stiff enough vibration can be a secondary issue. Damping does not change the resonant freq it means the vibration instead of taking 50 cycles to stop will take 5 cycles to stop as an example... Are you going to stress relief after welding? Peter



  3. #3
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    I have the model above fixed by the bolt holes in the base plate, I grabbed the cylinders that the bolts clamp through and used them as the fixture surface, so the above simulation assumes that there is no support under the center of the column and that it's just "pinned' to a completely inert base via the 8 1" bolt holes. Currently, it tells me that the column will resonate at 247.75hz, which is MUCH lower than desired. That's excitable by a 4 flute end mill at only 61RPM...and the vibration from any 60Hz transformers bolted to the frame will tend to excite it.

    I'm trying to find out if the result I have obtained by fixturing the part in the way I have, and running the simulation as a frequency analysis, will give me the data I need to determine how and where to add damping material or bracing to get the lowest vibration I can feasibly get for a C-frame mill.

    Damping *does* change the resonant frequency, because it will add mass. Any damping material added will add mass and thus lower any resonant frequency. Ideally, the amount of damping factor added compares favorably with the lowering of the resonance.

    I'm laying out the hole pattern and drilling the mounting pad for the tool changer this evening, and will do the ballscrew mounts probably later this week. They'll be welded to the structure, and I am looking into adding some 2" diameter steel tubing welded into the middle of the column to keep the 10" wide faces from vibrating, as the thing rings like a bell (as you'd expect from an open-ended tube) right now. I think two steel tubes, between the 10" wide faces, will help to control vibration of the wide faces and also provide an air circulation path to prevent coolant mist from condensing under the way covers, and it'll open up a wire path out of the ballscrew valley.



  4. #4
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - "Damping material" does add mass but that's not how it works. Stiffening something to restrain or change a vibration is a different tactic to adding dampening material. Your first order approach is to make it as stiff as possible. Since you have selected steel this is a relative lively material. Adding "structural mass" may not be the best solution. See page 9 of the PMD doc. Peter

    Attached Files Attached Files


  5. #5
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - I made a 100x75x10 thick 600mm long cantilever. Two in the model, one aluminium one steel. So the rigidity and weight are quite different. But the first vib for the steel is 211htz and the first for the Al is 214htz. Close enough to be the same in terms of a machine element. Then the second mode for steel is 267htz and al is 271htz. The al is quite a bit lighter yet vibrates at a little bit higher so this is why the mass theory has some holes in it... al beam 5kg, st 14kg yet the al has a higher vibration freq.... Peter

    unless you do a transient dynamic analysis including damping factors its tough to figure out if the design is damper then another one. Does your version of Solidworks have transient solvers?

    Attached Thumbnails Attached Thumbnails Welded Steel Frame VMC-Interpreting simulation results-vib-1-jpg  


  6. #6
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Quote Originally Posted by peteeng View Post
    Hi Switch - I made a 100x75x10 thick 600mm long cantilever. Two in the model, one aluminium one steel. So the rigidity and weight are quite different. But the first vib for the steel is 211htz and the first for the Al is 214htz. Close enough to be the same in terms of a machine element. Then the second mode for steel is 267htz and al is 271htz. The al is quite a bit lighter yet vibrates at a little bit higher so this is why the mass theory has some holes in it... al beam 5kg, st 14kg yet the al has a higher vibration freq.... Peter

    unless you do a transient dynamic analysis including damping factors its tough to figure out if the design is damper then another one. Does your version of Solidworks have transient solvers?
    Peter, we are in agreement as to the effect of mass on resonant frequency...I do NOT want to add more mass than needed, (increase mass) as this lowers the resonant frequency, exactly as your model predicts. I do want to add damping, and I understand that this does mean adding mass-but completely filling the column with epoxy-granite mix adds a LOT of mass, and may not be needed to damp the vibrations effectively, if I can get the column stiff enough to push resonances high enough. It's also quite expensive, and does not appreciably increase rigidity of the machine. (young's modulus is *much* lower for EG than steel.)

    I believe it does have the transient solvers, however...I have very little experience in using them. The first time I've used the simulation environment was to create the frequency analysis in the first post.

    Do you know a reasonable way to simulate adding a damping material in SW simulation? If so, I think it would be an excellent way to determine where to add damping material. My other options include wrapping the column in a layer or two of commercial butyl rubber based damping tape, with a thick aluminum skin. (It's a decent constrained layer damper.)

    Another user pointed out that I missed a RPS to RPM conversion-currently a 4 flute end mill would excite the column shown here at 3716 RPM, not 61! Big difference. Still, this machine will operate at spindle speeds up to 10,000RPM, so I need to push that up as high as I can. Higher frequencies are also much easier to dampen effectively, and require less material to do so.



  7. #7
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    After spending a LOT of time on this today, I've come to the conclusion that nothing I add is going to increase the stiffness or move that first resonance frequency. I can push the 2nd mode and higher modes up considerably-but the first mode just doesn't budge.

    Currently, Mode Shape 1 claims a resonant frequency at 211hz, and a displacement of 0.48. I'm using an IPS dimensional environment-does this mean that the simulation expects the top end of my column to whip around 0.48 inches at resonance?



  8. #8
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - The displacement in a modal analysis is just a % of the size of the model to allow the mode to be visualised. It is not a real displacement. In a transient analysis you will pick a time scale say 10sec for the solver period. You will deflect the model via a displacement shape or a force/time shape say a square wave. At a certain time say 2secs the forcing shape is stopped and the structure will go from full resonance to zero resonance depending on either the damping of the materials (material property) or a general structure damping ratio. Welded and bolted steel structures use about 2% damping. The solver then solves the decrement over time. For a live structure you may need more then 10secs solver time...

    The butyl tapes used for instance on thin automotive panels won't work on thick machine structures. Constrained layer damping needs the two sides of the butyl to be of similar stiffness to work. If it were that easy then commercial machines would use the stuff. To do a transient damping analysis you model the steel, butyl and foil. Each material will have a damping ratio or a hysteresis table. The transient solver then figures out the damping of the structure. Getting damping factors for these sort of materials is not easy though... There are many welded steel mills on the forum and other places that are very successful. Work on increasing stiffness. What do you intend to machine? and are you aiming at the same stiffness as a commercial VMC? dampness is a secondary issue. Machine rigidity is your first touch point to get right. Commercial VMC machines range from 100-150N/0.001mm upward at the tool. This is called the static stiffness. If this is high enough then you will not need to concern yourself with damping. Peter

    I also disagree with adding EG to thick steel structures as it is not stiff enough to contribute and its an expensive way to add mass. Depending on how far you are into the design and committed to the build, consider laminated steel as a good answer. This provides very good constrained layer damping and good stiffness.

    I looked up the SW solvers. It has as I described three ways to model damping. 1) material by material 2) a global damping factor and 3) Rayleigh damping which needs various damping and material properties or test data to tune to. Its not an easy area of analysis as the required data is not easily available,...

    Last edited by peteeng; 04-11-2021 at 06:51 AM.


  9. #9
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    I spent some time today tapping on the welded column and pressing my hand against various places to see where the ringing is happening, and using a microphone and laptop to figure resonances. The most effective face to touch, to dampen the ringing caused by a single whack with a wooden hammer handle, was the front 10" wide face of the column between the two linear rail mounting stands. Specifically, grabbing the still-open top edge, right between the rail mounts. Didn't matter where I tapped it at, that's the spot I could reliably touch and kill the vibration.

    That got me thinking, about putting a 10mm plate inside the column, and squeezing a 10mm thick layer of rubber between the inside of that front face, and the additional 10mm plate. I could then press in a spacer or two spacers that would compress the rubber layer between the new plate, and the inside front face. I'd be adding a rubber damper against the face that responded the most to my quick experiment today. Alternatively, I could pour in a liquid rubber between the two plates, bonding them in place, and let the internal plate wobble around willy-nilly.

    But-back to the primary concern, static stiffness. I want to make the stiffest machine I can afford to build, as I will be using it for both steel and aluminum work. I am targeting a Haas-style "Office Mill", as they use the same size linear rail hardware that I already have, and I think that the stiffness of the linear rails/guides is going to be the limiting factor-as the 25mm rails only use M5 bolts to attach to the column, so the limiting factor for stiffness is going to be the M5 mounting hardware.

    So, the machine structure needs to be limited by the M5 bolts holding the rails on. If I can approach that, I'm satisfied that I will have a machine usable for the tasks it needs to do-which is proof designs and prints, and the occasional short run. (10-50 pieces) Most of the work will be handled at the main shop, on commercial machines. This is so I can avoid driving an hour just to run one part, proof the model and CAM, before sending it out to the customer for their machines to run.

    The work envelope of the machine is 13" Y travel, 24.5" X travel, and 13" Z travel, and I can rework one of my Z ballscrews to obtain 16" of Z travel by removing the 2nd half of the split-ballnut and re-balling it to be a preloaded ballnut. The spindle is a BT30 taper, 8000RPM capable belt-driven unit, and will have 2kW to 4kW available, right now it'll have 2kW as that's the size of motor with matching drive I have on hand. There should be plenty of room to upgrade to a 4kW motor and drive at a later date.



  10. #10
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - Machine frames excited by hammers and the resultant "Ringing" may not be an indication of how it will perform in reality. Ringing is sometimes called acoustic vibration. Its like those annoying vibrations heard in your car. In my last truck I had one that I called the canary. Could never track it down. But the canary did not decrease the performance of my car. Same as this face of the hollow section. Hollows do suffer from acoustic vibrations. The easiest way to fix that is to fill them with 300kg PU foam. Not hollow anymore and the fill is viscoelastic.... In the many many machines I have designed (with most of them not getting off the HD) the screws fixing the rails have not been the limiting factor. Stiffness has. Maybe you could put up some more images we have not seen any yet.... X24.5(612mm) Y13 (330mm) and Z13" (330mm) Is a reasonable size machine and your column sounds light compared to a HAAS. Here's a shot of the CM-1 column about what your machine is.....especially take note of the bottom. Being a cantilever most of the stiffness is needed in the bottom 30% of the span.... Using standard sections has many issues. Since you are a welder (?) and have access to machine shop I suggest you fabricate the column and parts vs using std sections. Sections have rolled edges which flex and present issues for welding. Std sections by definition are std so you end up adding bits and pieces to it to either gain geometry or gain stiffness. Square and rectangualr panels ring etc etc.... The least amount of parts that get you to the result is the best approach...Peter

    Attached Thumbnails Attached Thumbnails Welded Steel Frame VMC-Interpreting simulation results-cnc-1-jpg  
    Last edited by peteeng; 04-11-2021 at 05:30 PM.


  11. #11
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Here's a few shots of the column, from a few days ago. I have finished the weld around the base plate, and gotten the two rearmost reinforcing ribs welded in.









    And the spindle cartridge, one of the Aliexpress BT30 spindles. These things are incredible value for the money. The two I have, the internal taper has less runout than my tenths indicator will reliably pick up. The bearing seem fine, mine have NSK bearings fitted, and I have picked up an extra set here in the states just in case. They do have some issues I want to correct, but I am not sure how I want to fix them yet. (No air blast through the spindle to clear chips when changing tools is one, and the ATC drawbar is push-only, with no loading flange to prevent loading the bearings with the drawbar on a toolchange is the other.)



    The mounting block for the spindle-another weldment. The bore is 86mm currently, and will have to be taken out to 90mm for the spindle cartridge. The plates welded in here are all 19mm thick, except the mounting ring which is 25mm thick.







    The spindle mount block is as big as I can fit in my heat treat oven here at home, I'll have to take the rest of the weldments to the local heat treater for stress relief.



    Into the old wore-out bridgeport for initial cleaning up and squaring up after heat treatment:





    Here I'm just measuring to make sure I can line-bore the spindle sleeve on the lathe, as my mill can't bore a hole this size, this deep, in one set-up. The lathe can easily.



    Before the spindle block gets line-bored, it will have the mounting pattern drilled in the mounting face, and the tram screws drilled and fitted in place. After that it's a tap-tap-tap and a shim here and there to get it lined up on the boring table, then bore through to fit the spindle cartridge.

    I have not found many references on cartridge spindle mounting that specified how tight the bore tolerances are. Some makers have the bore fairly loose, and only touching at the belt drive end to support the belt load, and all the tramming and spindle loads are carried solely by the mounting flange. Some other makers have the bore a close slip fit for the whole length. I'm planning to go with a full-length close slip fit, and when installing the spindle use a smearing of anti-seize to help transfer spindle heat out of the cartridge and into the mounting block. I've even considered wrapping the OD of the sleeve (in the weldment) with some copper tube, to circulate coolant through before sending it to the coolant flush nozzles.



  12. #12
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1254
    Downloads
    1
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Those risers under the linear rails are a lot taller than I was expecting. Why so tall?

    Closing the end of the RHS tube will improve torsional stiffness. Torsional stiffness comes into play a lot for X axis direction cuts (on the typical C frame mill layout)

    7xCNC.com - CNC info for the minilathe (7x10, 7x12, 7x14, 7x16)


  13. #13
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1254
    Downloads
    1
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    I am interested to hear more about the spindle cartridge over time etc. I will one day get around to my mill build and am intending to get a cartridge spindle.

    7xCNC.com - CNC info for the minilathe (7x10, 7x12, 7x14, 7x16)


  14. #14
    Member
    Join Date
    Feb 2006
    Location
    USA
    Posts
    116
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    looks great! I am interested in how it performs. I think there are a lot of methods available to damp the structure. and it looks like it will have good stiffness for accuracy in cut as well as positional deformation. what kind of stiffness are you shooting for? are you looking for a .001" deviation at 1000lbs of cutting force?

    Ray,
    Life is a choice, death is choice poorly made.


  15. #15
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - your links have stopped working. Can other lookers open these links? Switch can you use the thread tool please to include images, then they will be permanent... Peter see go advanced below / then manage attachments...



  16. #16
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Peter-the links are working fine for me. They're direct links to my image hosting, and CNCZone complains that the images are too large if I try to reference them locally. They shouldn't be links, they should be embedded directly into the posts.

    The risers are tall because they have to clear the 32mm ballscrews. They're C1 class precision, and I got 12 of them, new in the the original boxes, sealed in the bags, for 75$. NTN brand, double-nut with shims for preload, 32mm diameter and 10mm pitch. Sold off most of them, to buy the X axis screw, which is a Kuroda C1 precision 25mm, 10mm pitch...but 56" long and will have to be cut down to fit.



    https://i.imgur.com/tvJ5KCX.jpg



  17. #17
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    I can't edit the posts, so I will re-link the images here directly, Peter.

    This is a few views of the column:

    https://i.imgur.com/OyXapvc.jpg

    https://i.imgur.com/4eCtyBh.jpg

    https://i.imgur.com/RGiMZSW.jpg

    Here's a few shots of the spindle mounting block:

    https://i.imgur.com/RV8QzSw.jpg

    https://i.imgur.com/j7nBEPm.jpg

    https://i.imgur.com/jUC5LOn.jpg

    https://i.imgur.com/IXd7mKm.jpg

    Getting the initial clean-up done on the spindle mounting block weldment:

    https://i.imgur.com/dB2iBoS.jpg

    https://i.imgur.com/dqtWpue.jpg

    Setting up the mounting block to measure and be sure I can line-bore it on the lathe:

    https://i.imgur.com/OYoe7Ch.jpg

    These should be here for quite a lot longer, Peter.



  18. #18
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Got some more data here.

    Added some 3/8" gussets that will clear the ballscrew, and filled the column with epoxy-granite fill, using an 80:20 ratio of aggregate to epoxy. I used the parameters from the attached paper to generate the material in Solidworks.

    I then did a static load simulation, again, anchored only by the 8 1" diameter bolts that will hold it to the base, and applied a 1000ft-lb torque to the linear rail mounting faces, centered around the centerline of the main column.

    Deflection is quite good, I think.

    Welded Steel Frame VMC-Interpreting simulation results-1000ft-lbs-torque-column-jpg

    Welded Steel Frame VMC-Interpreting simulation results-1000ft-lbs-torque-2-jpg

    Attached Files Attached Files


  19. #19
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    2834
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Hi Switch - Interesting article. Their material only gets to <14GPa flexural stiffness. Its nearly not worth using at that stiffness and tensile is only 6.5GPa which is only double that of the epoxy... If this is the trend with quartz then I wouldn't use it. UHPC starts at 30GPa and is much cheaper then epoxy... Until you model the machine from tool to bed and then get its static stiffness its hard to tell how good your machine rigidity is... The Z axis is the stiffness killer.. Peter



  20. #20
    Member
    Join Date
    Mar 2006
    Location
    US
    Posts
    37
    Downloads
    0
    Uploads
    0

    Default Re: Welded Steel Frame VMC-Interpreting simulation results

    Peter...this IS the Z axis. There is a reason I'm working on it first.

    The material being used is "red granite", not quartz, in the linked article.

    Take note also, the simulation shown above is *with* the "nearly not worth using" material, and it's an order of magnitude better with the fill than without.

    UHPC is still water-bound, and will always and forever be prone to rusting a steel component in contact with it, and will not maintain stability over humidty and temperature changes due to this water content. Non-water-bound UHPC is just...epoxy granite with fancy aggregate.



Page 1 of 2 12 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

Welded Steel Frame VMC-Interpreting simulation results

Welded Steel Frame VMC-Interpreting simulation results

Welded Steel Frame VMC-Interpreting simulation results