What's new
What's new

Testing Camelback and Prism Straight-Edge sag and Vibration Stress Relieving.

Richard King

Diamond
Joined
Jul 12, 2005
Location
Cottage Grove, MN 55016
The last 2 days of my current 2 week scraping class I am teaching at a USA Defense Plant called DLA in Mechanicsburg PA I am going to have their Quality Control Lab test a 36" Brown and Sharp and Challenge Camelback that we are now scraping for flex. I am going to have their QC lab document the results for those who doubt me.

Yesterday we located some camelback Straight-edges in a cabinet that were hanging. The QC lab foreman said they had not been used in 5 years. I had my 3 best students start to scrape them. After we got about 15 PPI on them and hinging at 30%. I had them hang them up from a nylon strap on a gantry crane so they hung vertically and hit them with a soft blow hammer all over; the ribs, the spine, the flat scraped side. At first they were gun shy and I told them to hit them harder so they could hear the ring and feel them vibrate. To their surprise all 3 moved the hinge point toward the ends about 10% from the ends. That was under .001" but they MOVED.

The QC Lab checked and they had been in inventory for over 50 years, so they were well aged. I will add photo's and the QC lab report as I get them. The plan is to set the Straight-Edges on the same size gage blocks and set a weight on the center of the SE's and see how much they sag or bend. They will do the tests inside they temperature controlled lab on their Grade AA granite plate. We will test them as soon as we set the weight on the center and then leave the weight on over night.

I was taught to ring or vibration stress relieve straight edges buy my Journeyman Dad who was taught by German Journeymen to Ring dropped or fresh casting and machined SE's and in all my years they all change.

We do this because if we didn't, next time we used them and it gets bumped a little hard when setting it down on a way it "COULD" move and we would get a false reading. It is standard practice and not difficult to do.
 

Attachments

  • 20230215_151809 (1).jpg
    20230215_151809 (1).jpg
    1.9 MB · Views: 40
Last edited:
Other things we're discovered while scraping the 30 Starrett 199 levels and Precision Blade Squares. On the levels we had to wear cotton gloves or pick them up with a white cotton rag or the heat of their hands could change the hinge or spin test when scraping and testing. We took off the plastic top where your are suppose to hold them and the level vial so we did not damage it. With the bare cast iron if we did not insulate the castings from the heat of our hands they would heat up super fast.

We used a infa-red hear gun to test the temp difference. The 2 ends would read 72 degree's F and where we would hold the center with out the gloves it would read 70 degrees F. If we insulated the casting with the cloth the whole length read 72 deg. F and that's how we had to check them. I knew hands transfer heat, but never thought holding them in the center for less then 1 or 2 minutes as the student walked from their work bench to the granite could change the pivot point. This old dog learned a new trick. :-) I will attach some photo's tomorrow

They had several blade squares that were out of spec. and were sitting in the reject shelf. We used their Square-Master to check them before and after we scraped them.
 

Attachments

  • 20230209_151634 (1).jpg
    20230209_151634 (1).jpg
    7.3 MB · Views: 61
Last edited:
I'm tired, so no long winded explanation as some like to write - Long 2 weeks of class. Today 2 of the 3 camelbacks were finished and the prism SE was 80% done. The QC guy was anxious to test them after we told him about our experiment we wanted to do and the 36" we took it into the QC room and they tested it. He used a push and pull tester (He called it something, but I forgot what it's called) on the push down. It was in Kilograms - Not my specialty...metric..lol He set them on his granite plate on the same height parallels and we first set the 36" camel back on them and he pushed down on the center and we had a .00005" indicator under the center bottom. He pushed down 20 kg (approx. 40 pounds) and the camel back bent or sagged.0001".

He then set the prism SE on the parallels, with the indicator under the center. Pressed the SE down with the gage and it sagged .001" He then slid his finger under it and lifted the SE .0002" with ease. It was sagging with it's own weight too. I asked him to write down what he did and I got the look as he really didn't care.
 

Attachments

  • 20230216_094051.jpg
    20230216_094051.jpg
    4 MB · Views: 40
  • 20230216_095110.jpg
    20230216_095110.jpg
    1.5 MB · Views: 40
  • 20230216_095116.jpg
    20230216_095116.jpg
    1.4 MB · Views: 39
To clarify: I could have asked him to press down 5 or10 pounds and it would have bent He picked the pressure he pushed down. I'm trying to inform people not to trust a prism SE compared to a camelback. That prism SE was thicker then some thin ones on the market or home made. One more day and tomorrow I will test it with a lighter pressure. It bent .0002" from it's own weight.
 
I was showing the students how I straighten gibs using a C-Clamp and Indicator, with the gib sitting on 2 same size steel blocks on a steel bench, use a C-Clamp like their machine. and they brought over the gib straightening machine. You check the gib with feeler gages and then put it on the blocks and turn the screw. I mounted the mag base and indicator behind the machine and using the indicator screw it up and down until the gib moves the amount of the feeler gage.
 

Attachments

  • 20230216_110821.jpg
    20230216_110821.jpg
    3 MB · Views: 44
  • 20230216_110826.jpg
    20230216_110826.jpg
    3.1 MB · Views: 44
  • 20230216_110834.jpg
    20230216_110834.jpg
    3.2 MB · Views: 43
Rich,

Comments:

I think you did some real good by making careful measured observations. The observations concerning ringing are fairly compelling. It makes you wonder if there could be a super thin layer of fused/welded crystals left behind by the scraper pushing the metal off and, as it does so, melting a micro-thin skin of thin fragile metal. (I am recalling those micro-videos of cutting tools plowing through metal showing molten metal at the interface of the tool and the metal). Could the vibration crack that skin? A way to test it might be to subject a recently scraped SE to very light etch using acid and testing it before and after etching. The best test along this line might be to do it on a parallel with both sides scraped very flat and then one side lightly etched and if the unetched surface changed its hinge point after etching that would be very suggestive of the skin formation.

Another question that occurs to me is if scraping induces a stress that can be relieved with ringing, and the straight edge is necessarily re-scraped to eliminate the deflection caused by ringing, when do you stop? Or is the stress induced by finish scraping compared to roughing so slight as to be imperceptible? Successive etch tests on a parallel could speak to that question by scraping, testing a surface, ringing it, and re scraping and re etching.

With respect to prisms sagging more under their own weight than a camelback I would say it is pretty clear the design of the camelback and of the prism will strongly affect their rigidity and that result would be expected. (I think we all know there is a reason engineers design their bridges to look more like a camelback than a flat bar of metal.) I will be testing my cored-out 18" and 26" prism designs in the near future. As this cored-out design is an intermediate between a flat bar and an arched-bridge design and more like an I-beam you would expect it to flex less under its own weight than a solid bar but more than the bridge. As it turns out, in actual practical use neither the camelback nor the prism are suspended in that way. So, either can be "trusted" for straightness when used as intended. It seems like stability is the key factor.

I have been observing a new thermally stress relieved 36" Featherweight Camelback suspended between 2-4-6 blocks for almost a month. It has been undisturbed in that time and has shown stability---no progressive sag. The ten millionths indicating gage reached thermal equilibrium after a day or so and has remained on its same reading without varying more than an interpolated 1-2 millionths in that time. So progressive sagging seems not to be occurring and progressive sag due to creep (taking a set) is not expected. It, like every structure in the world, sags under its own weight.

I will be doing some more testing of a slab of cast iron subjecting it to a fairly high load for a prolonged period of time to see if any "set" can be induced.

With respect to distortion of the 18" level in just a minute or two of hand contact, I am surprised it was not worse. In my own testing it took only 90 seconds to distort my angle plate a full tenth at room temp and around 15 seconds at 45 degrees room temp. At 45 degrees holding my camelback with a bare hand on the mid arch caused .0001 deflection in 10 or 15 seconds.

Interesting stuff. Most surprising is the ringing test. Good for you. If it were possible to do the etching test that might lend more information as to what is going on.

Denis
 
I did another test on a thinner and 48" prism SE and the same camelback. The same QC guy di the pressing and the camelback stayed the same .0001". On the thinner prism we did a gradual 5 10, 15 and 20 kg again it gradually went from .0002, .0006, .0015 and finally .0025" So TGtool they bend differently by length and design. I tried to take pictures as it went, but this si as good as I could. I am not going to submit this as scientific fact paper, but I showed those who care prisms bend compared to camelbacks. I knew this since I was an apprentice.
 

Attachments

  • 20230217_083649.jpg
    20230217_083649.jpg
    2.6 MB · Views: 17
  • 20230217_084006.jpg
    20230217_084006.jpg
    3.2 MB · Views: 17
  • 20230217_084009.jpg
    20230217_084009.jpg
    3.4 MB · Views: 17
  • 20230217_084019.jpg
    20230217_084019.jpg
    3.3 MB · Views: 16
  • 20230217_084026.jpg
    20230217_084026.jpg
    3.2 MB · Views: 16
  • 20230217_084032.jpg
    20230217_084032.jpg
    3.1 MB · Views: 16
I absolutely agree that the prism will bend more than a camelback design under either its own weight or with any amount of load. That's not even a question.

My question is more like, "How much more is a prism sagging when supported at the Airy points?" By definition those are the support points that produce the minimum deflection. So if you're scraping a surface and it's not yet completely flat but has a slight hole in the middle, you put your reference down (prism or camelback) and the reference is now supported somewhere in the region of the Airy points. In THAT situation, how much more does the prism sag in the middle compared to the camelback? That's more nearly the real world situation than the reference supported on its ends, weighed in the middle on not. Is that enough to make a material difference, or is the difference in sag at that point almost lost in the noise of evaluating flatness and reading the spots plus whatever vagaries of holding, pressing or sliding the straightedge and any unevenness of the spotting compound?
 
I absolutely agree that the prism will bend more than a camelback design under either its own weight or with any amount of load. That's not even a question.

My question is more like, "How much more is a prism sagging when supported at the Airy points?" By definition those are the support points that produce the minimum deflection. So if you're scraping a surface and it's not yet completely flat but has a slight hole in the middle, you put your reference down (prism or camelback) and the reference is now supported somewhere in the region of the Airy points. In THAT situation, how much more does the prism sag in the middle compared to the camelback? That's more nearly the real world situation than the reference supported on its ends, weighed in the middle on not. Is that enough to make a material difference, or is the difference in sag at that point almost lost in the noise of evaluating flatness and reading the spots plus whatever vagaries of holding, pressing or sliding the straightedge and any unevenness of the spotting compound?
We use these tools for a practical purpose. if you are scraping a worn way that is concave and were to use a prism style straight edge you still have somewhat of a convex way surface. The camel back is a beam, it sags less by quite a bit.
The airy points don't really matter when your rough scraping in a way that is a few thou out. It does matter when you are finishing and need to check hinge. If you've scraped a way surface to a sagged prism you will always have a convex way. Use the camel back's inherent rigidity to establish a flat plane.

Prisms can be good once you have established a flat plane and all of your other geometry is known and well understood. If your way has been roughed in flat then why not use a prism?
 
I too think it's not a question of whether the sag will happen, but by how much. I think without question the camelback straight edge will be stiffer than a prism and that it's going to be better to use a camelback wherever possible, but sometimes that's not easy to fit somewhere or perhaps just out of reach financially for the guy working on his own stuff. For a longer surface I probably would not trust a prism unless I could double check it against a camelback. For a shorter surface I don't think it's a big problem, as the prism won't sag too much under its own weight at shorter lengths. They're never going to be weight loaded at center in use, so the only thing that really proves is that the camelback is stiffer, which I'm not sure anyone would argue against.
 
I too think it's not a question of whether the sag will happen, but by how much. I think without question the camelback straight edge will be stiffer than a prism and that it's going to be better to use a camelback wherever possible, but sometimes that's not easy to fit somewhere or perhaps just out of reach financially for the guy working on his own stuff. For a longer surface I probably would not trust a prism unless I could double check it against a camelback. For a shorter surface I don't think it's a big problem, as the prism won't sag too much under its own weight at shorter lengths. They're never going to be weight loaded at center in use, so the only thing that really proves is that the camelback is stiffer, which I'm not sure anyone would argue against.
Let's just say for the sake of discussion see if it really mattered that your reference surface whether it is a prism or camelback bows when end- suspended? What would happen to a scraped surface made using such a reference surface? (That really should read "ANY" reference surface)

Assume the reference surface sagged a tenth when suspended but was scraped to match a flat plate. Could the work surface be scraped to match with a .0001 hollow in it that would just match the reference bow of the end-suspended reference. That won't happen as the reference, in contact with a work surface, is suspended along its entire length at closely spaced (40PPI?) intervals as if floating on a thin spotted film of ink. It is like it is on closely spaced springs. (I sort of recall this conversation from 5 years ago or so.) It will straighten as it is better and better supported as scraping progresses.

Or imagine you are scraping a work surface that is slightly hollowed out and is corrugated, maybe from milling. Still the "noodley" work surface will ink the high spots which will be scraped away in repeated passes and the reference will gradually be better and better supported along its length at shorter and shorter intervals until it relaxes to its straight condition as originally scraped and the work surface will become flatter to match.

So, that is my take on why the question of end-suspended stiffness is essentially moot in practice. Still it comes up for discussion and it is good to have a general understanding of the relative stiffness of reference surfaces. Maybe there is a case when it actually matters?

Denis
 
If the machine wears like 90% of all ways wears, it's high on the ends and low in the middle. Many rookies buy or make prism because they are cheaper or easy to make to rub a dovetail. They hold it in the middle and their hands push down and it looks like it's touching in the middle. Thats why we make a camelback with a 45 or 50 degree angle on one side. You can buy a new designed SE, but their are so many prism SE's out there, not everyone is going to buy one or a camelback with and angle. I AM Warning them to be careful not to trust the reading or bluing.

We all are saying the same thing to some extent aren't we? When I use a prismI on a long part like the base of a centerless grinder I use both.

Next.
 
I think one case where it definitely matters is when you're testing a short surface with a straightedge that's much longer. A prism will sag at the ends making the work surface high in the middle when the contact pattern looks perfect.
 
Concerning the question of grey iron sagging if stored in a strained condition, I just completed a 1 month trial of placing on of my raw stress-relieved 36” camelback castings on 2-4-6 blocks placed one at each end and then positioning a sensitive electronic lvdt at its mid point. In that time there was no progressive change. It did not vary from its initial position more than +/- 10 millionths and those variations were not perisistent. In other words, it’s measured position (sag) did vary up and down by at most 10 millionths and usually far less than that. The 3 point suspension shown above in this thread proved very stable. I repeatedly stressed the suspension and the needle consistently came back to its pre-stressed position, Those needle fluctuations were random and most likely were related to temperature variation in the room. The LVDT was left powered on the entire time to avoid heating cooling drift due to its power consumption. After just a day or so of being powered on it seemed to reach equilibrium. But when first powered on you could see it drift maybe 20 millionths over a few hours while its various components reached their steady state temperatures.

Next up I will be subjecting a non-stress-relieved slab of cast iron to significant strain and will be using a better comparative measuring system on it to see if creep can be introduced that way.

I am working these tests in between casting, machining and packaging straight edges and making some furnace upgrades and a new core box for my 18” prism—no rest for the wicked as they say. ( A machined 18 goes to Thailand tomorrow. A 36 and 48 are trundling across Minnesota right now headed to NY. A 36 goes to Georgia tomorrow as well.)

Denis
 








 
Back
Top