Flattening a table larger than your straight edge - Page 2
Close
Login to Your Account
Page 2 of 2 FirstFirst 12
Results 21 to 24 of 24
  1. #21
    Join Date
    Feb 2010
    Location
    ch
    Posts
    2,501
    Post Thanks / Like
    Likes (Given)
    215
    Likes (Received)
    329

    Default

    interesting idea. can you give an example of the "heating elements" please? they must be insulated, right? not just the plain spiral wire from a regular heater. or are you talking about a kitchen oven? it never occurred to me you could straighten those.

  2. #22
    Join Date
    Aug 2017
    Country
    UNITED STATES
    State/Province
    Florida
    Posts
    44
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    0

    Default

    Quote Originally Posted by johansen View Post
    what i would do with more time than money:

    assuming the bed is convex, as most are due to peening stresses:

    machine 5 rods to the same length, from 3/4 or 1" steel. they need to be about 9" long. drill and tap one end of them for whatever you use for T nuts, tighten them up against 5 ground washers, such that they are all the same height on your table. one in the center, 2 at each end. if you have an even number of T slots, you need 6 posts. two in the center off set from each other such that they match the same straight edge.

    build a straight edge long enough to measure the diagonal:

    you only have to measure ~50 inches so i would marginally suggest the 2" square steel square tubing with pre-existing holes, the stuff they use for road signs.

    make 6, 1/2-20 bolts that are all exactly.. say 4 inches long, with the ends ground square to the thread. you can grind them by eye if you have a v block to rotate them in next to a grinding wheel. .0005" tolerance on the length is probably good enough for you.


    find the holes in the beam that are closest to the diagonal of the posts you bolted to the bed, or move the diagonals to match the straight edge. if you have an even number of T slots you need to move the middle post in the straight edge towards one end, and offset the two middle posts in the bed to match the straight edge across both diagonals. i tried drawing an ascii image but it failed miserably, i think you can figure it out.


    pre-set the 1/2-20 bolts to the same height off the steel tubing, one at each end, one in the middle. tighten the nuts but not very tight, weld the nuts to each side of the beam with a tiny tack weld. if you drill and tap regular square steel, use a lock washer, or a nylock nut.. you want friction but not the hassle of jam nuts where you have to lossen them to rotate the bolt.

    lay the two beams next to each other on a reasonably flat surface (a good sheet of plywood is good enough, or use your surface plate, and adjust the center bolts of each beam such that flipping either of the beams over doesn't result in much difference. you now have two identical straight edges each with 3 points in a straight line. you can add more bolts to get more points, but the effort in adjusting them grows significantly.


    now measure the diagonal of your bed. using shims measure how much its curved. you will want to average the measurements using both sides of both straight edges that you made. or, don't worry about it and just use one of them and note which side you are using.


    now take the bed off the mill, support it on three points. one in the middle of one end, and the other end at two corners. i would weld 3 posts, about 2 feet high, to a somewhat rigid frame, but i don't think it matters. you want something that won't get knocked over.... then set the bed on those posts.

    get 2 oven heating elements (3500 watts each typically) and straighten them out, they are typically as long as 8 feet and you should be able to get 2 of them to fit down in the T slots in a U configuration. run them on 120 vac not 240, each will generate a fourth of their nominal wattage.

    wrap the bed in 3 or 4 layers of regular r-13 fiberglass insulation, and heat it up, you should be able to reach 1200F with 1750 watts of heat. I was able to hit 1000F on an 8 foot long, 9 inch by 9 inch square beam wrapped up in 2 layers of shitty fiberglass, but with 3500 watts of heat.

    you should have already noticed the bed sag, under its own weight, how much did it sag, keep that in mind....

    as it gets hotter it will sag more. and more quickly. it will take a long time to get up to temperature, 6 hours perhaps or more. you will want to check and recheck how much its sagging, and also how quickly. as it gets up past 1000F you are in the danger territory where you want to cut the power and unwrap the outerlayer of fiberglass insulation to let it cool off quicker.

    you could try a couple cycles. take it up to 800F and let it cool and see if it sagged any.. i am curious to know.


    instead of using shims and the straight edge you made to verify if its straight as you are heating it.. you can can use just any old piece of steel and rest it on the diagonals (with a strut to stabilize it against the other corner) and use a dial indicator to watch how much the middle post is sagging under the heat. if you trust your measurements with the straight edge you made, just note how much the table sagged under its own weight and add that to the amount you need to wait for the table to sag as the stress is removed. yes, the beam you've got your dial indicator clamped to is going to get warm might bend a bit, but that's why i suggested make the posts 9 inches long and use 3 or 4 layers of fiberglass.. or set up a fan to blow air on it to cool it off. or use a piece of aluminum which has better thermal conductivity

    the only reason to use the straight edge you made to measure how much it is sagging under its own weight is if you're paranoid that you're going to end up with a table bent in a helix.. personally i am not worried about that.. i think that the stress its under, under its own weight will dominate and you don't need to worry about it bending into a helix.

    now there is a problem with applying all the stress to the middle of the bed.. this isn't ideal. but, in the real world where most of the peening stresses were probably applied to the middle 2/3rds of the bed it is probably a good compromise.

    moving the 3 points you rest the bed on.. moving them in just a little bit like say 5 inches will make a pretty significant difference to how much stress the center of the bed is under.

    on a 50" bed moving the points in 12" from the ends will reduce the stress in the middle of the bed to zero and you will actaully droop both the middle and the ends an equal amount.
    Interesting idea. When you mentioned bolting the 1/2-20 bolts to the steel tube you said tighten the nuts, am I securing bolts to the through holes in the steel tube and just resting them on top of the posts or are they threaded into the posts as well? Why not use a ground shoulder bolt so I have an exact height measurement for each bolt?

    Also that last point you mentioned about moving the pins 12" in, are you referring to bessel points?

    Sent from my SM-N950U using Tapatalk

  3. #23
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,257
    Post Thanks / Like
    Likes (Given)
    261
    Likes (Received)
    297

    Default

    Quote Originally Posted by marino1310 View Post
    Interesting idea. When you mentioned bolting the 1/2-20 bolts to the steel tube you said tighten the nuts, am I securing bolts to the through holes in the steel tube and just resting them on top of the posts or are they threaded into the posts as well? Why not use a ground shoulder bolt so I have an exact height measurement for each bolt?

    Also that last point you mentioned about moving the pins 12" in, are you referring to bessel points?

    Sent from my SM-N950U using Tapatalk
    so what i'm talking about is manufacturing a straight edge.

    The minimum amount of information you need to make a straight edge is 3 identical length rods, two reference points and a movable third point.

    or: 3 beams, each with 2 fixed points on the ends, and you adjust the third point in the middle of the beam, of each of the 3 beams. this is slower than the method i've provided, but its more accurate because you don't have to start out with 3 rods of identical length.

    so you can simplify your life a bit here with just 2 threaded rods the same length as 4 non threaded rods.

    tac weld the two outer rods 50" apart or whatever you need, to the ends of the two beams you want them parallel to each other to within reasonable limits otherwise when you flip the beam over you wont get good contact area on them. while repeatedly flipping the beams over, move the two threaded rods in the middle of the beam such that they match each other. you now have 2 beams, each with 3 surfaces on each side of them, in a straight line. you do need to get the rods flat and coplaner within reasonable limits so you get good contact on the full surface of the 1/2" diameter rod.

    you then use this beam with its 3 surfaces to verify the twist of your table, and the convex/concavity of it, you will want to average the data from both sides of both beams if you want to get it close to zero.



    the posts you need to bolt to the bed are required because you can't measure the twist of the bed directly when its 900F

    the posts need to be the same diameter and the same legnth and the same material because one end of them is going to be 900F and the other end maybe 500F and you want them to all grow the same amount.

    the posts are just to transfer the information from the surface of the table up through the insulation so you can rest your straight edge across the diagonals of the post and watch how much the table twists or sags under the heat.


    Also that last point you mentioned about moving the pins 12" in, are you referring to bessel points?
    yeah. take a look at how severely the stress in the table drops off as the support points are moved in.

    Airy points - Wikipedia

  4. Likes dcsipo liked this post
  5. #24
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,257
    Post Thanks / Like
    Likes (Given)
    261
    Likes (Received)
    297

    Default

    Quote Originally Posted by dian View Post
    interesting idea. can you give an example of the "heating elements" please? they must be insulated, right? not just the plain spiral wire from a regular heater. or are you talking about a kitchen oven? it never occurred to me you could straighten those.
    i have straightened out the 2600 watt, spiral range elements but that is a tremendous pain in the ass. the 1800 watt small range elements are even tighter wound.

    the oven heating elements have much more generous bend radius.


Tags for this Thread

Bookmarks

Posting Permissions

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