John, love the video. Long time watcher, first time commenter. Im not a machinist so you’re always way over my head But as someone who has worked in force/load cell calibration for years, you need to be mindful that the load cell needs to be loaded at the points where it is threaded. Each side of that s-cell is a lever and as it deforms by design to give the signal that the strain gauges pick up and convert to force readings the fulcrum point changes when force is applied to the flat surface and applied load changes. Using the threaded holes and a bolt through them ensures that the fulcrum point is in the same place as when it was calibrated, increasing your accuracy.
This is the type of video that really makes you stop and think how many different variables a programmer has to consider when parts could be ran by multiple operators. Keep up the good work :)
Based on your setup, there is a very simple answer as to why clamping area on smooth jaws did not affect the force required to move the part (hint, it has nothing to do with pressure). The force applied by the jaws acts 90 degrees (perpendicular) to the "pushing" force you are applying. Therefore the 2 different forces act completely independently of each other on part (think drawing the forces as X and Y directions on a graph). Then what keeps the part from immediately moving? Friction! The friction force between the jaw face and part is what keeps the part in place. If we knock the dust off our handy engineering textbook, the formula for calculating friction force is: Friction=(perpendicular force)X(coefficient of friction). Notice how the formula does not contain anything about surface area, that is because as counter intuitive as it is, friction is completely independent of surface area and there is our answer. For fun, we can actually calculate the coefficient of friction between your vice jaws and the part using the numbers from your experiment. You clamped the part with about 3300lbs of force and it took about 550 lbs to move the block so if we plug those number in we get: 530=3300X(coefficient of friction) so the coefficient equals about 0.17! However, this only holds true for the very specific setup you created here. You still want to hold on to as much area as you can because cutting forces on the part never act just in a singular direction like your test setup. The force most likely to dislodge your part during machining is from torque or tipping force from the tool pushing on the edge of the part. To best counter this, you'd want to reduce the lever arm from which the torque is applied by leaving the minimum amount of stock exposed.
Very well put Mech engineer. Was gonna basically post the same thing along with my post above about why the variation when using the torque wrench but luckily having seen your post that already explained it, I didn't have to take the time.
Considering the other force directions, do you think it can be defined the minimum area required for an specific part to be in contact with the smooth jaws in order to stay still and not move?
Friction force isn't a function of the Area: F =µ*N (µ: Friction coefficient, N: Normal force). I guess Talon jaws add plastic deformation to the equation which isn't very effective at high forces as Aluminium is quite soft and therefore deforms easily
I was wondering about the same thing. Talon jaws grip really well, but aluminium just is too soft to hold on. Now you'd need to test smooth surfaces vs gripped jaws with different materials :)
John, having tested several torque wrenches a few years ago, I found that the number one cause for error is how you use it when it comes to clicker type torque wrenches. Clicker type wrenches are not good for when you want repeatable accurate torque because of the way they work. They are really more of situations when speed is of more of a concern than accuracy and and being repeatable. If one does not instantly let of pressure the moment the click happens, you will get a bit more torque applied than intended but this is really a relatively small amount of induced error in most cases. Where the real problem with them is though is the way you pull on the handle. If you are not pulling PERFECTLY perpendicular to the axis of rotation of the fastener and within a few degrees of perpendicular to the beam of the wrench, you will get more torque before the click occurs than intended. This is because of the friction induced in the hinge that allows the beam of the wrench to slightly deflect when the detent releases. Basically the detent doesn't roll at the set spring tension because friction of the pivot is preventing the detent from seeing the real force being applied.... Essentially, friction at the pivot is acting as a path for force to bypass the detent. If not quite following what I mean by perfectly perpendicular to the axis of rotation, think of it in the sense of this video and using the torque wrench to tighten the vise. if you are applying any force to the handle that is either towards you (the operator) or away from you (basically force that is in the same direction as the fastener, you are adding friction to the pivot which means that the detent is not seeing the real force, but rather a force that is less than what you are really applying. For illustration purposes, have a look at this illustration. en.wikipedia.org/wiki/Torque_wrench#/media/File:Direct_pawl_clicker_torque_concept.png if you are applying any force that is outside of perpendicular axis of rotation of the fasten, that force is creating friction on the pivot pin which in effect is causing the assembly to act more like a rigid member. Until enough force is applied to overcome this friction, the detent (red square peice) never sees the real force. So lets say that you are generating enough friction that it takes 10lbs of force to overcome the friction, that is 10lbs more force that is being used to create torque on the fasten.. Essentially, its like you added 10lbs of force the the spring Beam type torque wrenches are best if you want repeatable and accurate (if its a quality and accurately calibrated wrench) torque values but at the expense of speed. Beam type wrench are unaffected by forces that are outside of being perpendicular to the axis of rotation (as long as the beam needle is still floating). This is assuming that you cant use the both stretch measurement method (which is really is the most accurate method obviously). Hope this helps for anyone to understand why / how there could be so much variation from person to person and even for two different values from the same person.
For the drop off in holding power for the griping jaws you may be hitting the yield strength of the aluminum part itself due to the side pressure acting over a smaller area
Hi John, Have you tried or heard of "Electro Spark Deposition", you can have grip faces in steels, surface coated with a very hard textured/abrasive media, imagine gluing 40 to 100 grit abrasive cloth/paper to grip surfaces. Excellent in cutting through mill scale, and can be applied to any shape surface. Rocklanise, Surmet or Carbinite are the suppliers that come to mid. Great subject.
I think you'd want high clamping force, but low clamping pressure. That means grip as large an area as you can afford. The higher the pressure, the more you will damage the surface.
True, the clamping force is what is most important and will be what holds the part in place, pressure is bad in this case as it will lead to dings in the part.
Very interesting, for the torque wrench, I noticed each person used it differently, very differently, also the Min-Max delta, or range was massive. Like the data though!.
Here's an idea in case you decide to do another video: test how consistent the clutch settings on electric drills are. And/or try an impact wrench (electric or pneumatic) anf a couple of those extensions that are supposed to limit torque. Great info in this video, it is much appreciated!
I have learned much from your videos. Thank you. In a prior life I owned a company that built high performance engines, etc. for mostly Porsche's. I just wanted to make some observations. 1. I always use the same method to tighten bolts using a torque wrench. That is to first get the fastener barely tight so that the pull for the actual torque setting can be done in a single smooth stroke; stopping the click or beep. Also, the pull direction should be in an arc that is on the same plane as the rotation of the screw/bolt, ie., do not jiggle it forward or rearward. 2. Manufacturers almost always provide torque specs on the high side. 225 ft lbs on a pull stud is huge. I have seen the number used in some threads. I have a HAAS TL-1 lathe and HAAS TM-3P mill. The mill has a spindle holdfast included. If I put even close to 225 ft lbs on it, I am sure it would rip the unit off the machine. Not to mention that using a 1/2" torque wrench to do this would be outside af many folks strength. Similar to asking any but the most experienced machinists to measure anything at the .0001" level even using the proper mic. 3. My experience tells me that applying too much torque to a part is a much more prevalent and damaging problem than not enough. Using a 6" x 4" x 1/2" piece of 6061 T6 aluminum gripped 1/4" deep along the 6" side using smooth jaws, in a ORANGE vise, the workpiece will begin to deflect at about 45 ft lbs (on a calibrated torque wrench). At 100 ft lbs, the deflection at the centerline is more than .002". The worst part is that if you face cut that part, you would not even notice. If you measure all 4 corners you would;d not notice. If you measure the center line, you will notice. 4. You hit the nail on the head. Even with proper use and a quality torque wrench, you will get different results. If you include the torque wrenches that most people buy, results become inconsistent (the worst) and wrong. I am not sure if most folks return their torque wrench setting to the lowest setting after each use. 5. I know there are countless machinists out there that know what to torque their work pieces. Too bad that I have never seen a post anywhere like this: "I have a high quality torque wrench that was recently calibrated, an ORANGE smooth jaw vise, HAAS Mill and run at high speeds and feeds. Here are some example torque values I currently use:" Then, they could list several work pieces that they mill and include the vise torque values, metal type, size, gripping type and area, cutter info, speeds and feeds, etc. IF I HAD THESE REAL WORLD EXAMPLES, I COULD EASILY FIGURE THE REST OUT.
One thing about holding on a small amount of material, is that you will end up with deformation on your part, also if your part is up out of the vise more your tool will have a easier time rotating the part out rather than pushing it’s all about leverage.
This is really great work you're doing and sharing with the rest of us and I truly appreciate it, thank you. But I haven't noticed a lot of issues with parts moving sideways in the vise. It's typically pulled up. I imagine if you do the same series of tests but with a jack screw to apply upward force you'll see very different results more aligned with what we would expect. ( drill a clearance hole for the jack screw through the fixture to test super glue). Next you'll have to test what kind of forces are put on the part during cuts, climb vs conventional, slot milling vs side milling, rigid tapping, and of course how that changes with tool size. Keep up the great work looking into the science of machining.
The thing I found most impressive about using superglue to fixture a part is the tapes holding strength. Makes me want to see the differences between different types of tape.
Excellent and informative video. Being old school T&D guy. We used traditional vise handle with a pipe...if we needed heavy gripping. Wow were we wrong....
Hey John, check this out. I used to work for a Top Fuel Drag Racing Team that was sponsored at various times, by Snap-On, Matco, Cornwell Tools, and Travers Tool Company. There was a toolbox in the machine shop that had probavly 75 torque wrenches in it, high dollar Snap-On, Matco, etc. There was also a digital torque torque tester mounted on the wall by the box. Out of all of the torque wrenches, I believe only a total of 2 were accurate. Accepted practice was to set the wrench to your value needed, say 80 ft. lbs., and keep adjusting the torque wrench, until you hit the number required. So, even though the Snap-On (most expensive) cost a buttload of money, it was no more accurate than anything else.
What else may be overlooked is the actual mating surface geometry ( geometric tolerances) The modulus of elasticity of the vise is fairly constant, since the materials are similar. The part being held , aluminum , is less , and hence will distort elastically against the smooth mating face of the hardened steel jaw much sooner. The vise is just an irregularly shaped spring, and will deform as the applied load increases. Therefore, there are a lot of unseen forces that will alter the shape of the held part that just aren’t perceivable. You essentially are machining a squished part ( elastic deformation) , and the forced aren’t just in compression, but an increase in buckling as the contact area is less and less as the part is gripped higher in the vise. So, as already stated, the lowest pressure on the faces of the part is ideal, with a margin of safety. 1) keep part from moving or releasing during cut. 2) keep part from being too elastically , hopefully not plastically deformed while held during machining.
Thank you John. Test 2. Come up with some correlation for appropriate clamping force before a part will bow or distort. Subject to part thickness or cross-sectional area. I've always wanted to understand this better.
John....good start to understand how much work holding force you have using different methods and inputs. I would suggest when you try a new variable/process, do it more than once. This will help you see trends that you don't see with a sample size of 1. With enough measurements, you will start see the standard deviation of the process which will tell you how much variability (force) there is at confidence level. Or simply stated, do you have a robust work holding method. I would suggest looking at your work holding stiffness rather than ultimate clamping force. You don't necessarily care how much force the vise puts on the part as long as it holds it in place. However if are trying to do precise machining, you would want a stiff work holding method so it doesn't add to your tolerances.
Could you do similar tests with Mitee bite fixture clamps? It would be great to know a range of RDOCs and ADOCs with speeds and feeds to go with the different clamp sizes.
Great Video. But It would be more practical to know how much clamping force is needed? How hard to tighten your vice when using a 10mm endmill adaptive? Does calculating cutting force is enough or is vibration and dynamic load loosen up the part at this force?
On the torque wrench: buy a quality unit, they're not that expensive to get a good one. Make sure you pull perpendicularly to the handle from the hand grip and try not to induce twist. Come up on the desired setting very smoothly and slowly, and trust your instrument, no need to double click. Besides, if you do it right and try for a second click, it should not move the bolt at all as static friction is higher than the dynamic friction. Edit: I was further watching the video and saw on a few occasions where things were tightened up with somewhat outstretched arms resulting in the spanner jerking as it was tightened. In those cases use a longer wrench and hold it closer to your body if possible to try and get absolute smoothness when tightening
Cool experiment! So, first thing is that the amount of clamping load your vice jaws impart is a function of several things. The screw pitch, applied torque but also, and very importantly, the friction present within the vice assembly. If you want to substantially increase the repeatability of the clamping load the jaws impart then you need to do two things, first is to lubricate the vice screw and sliding surfaces. Second is to change the procedure you're using to clamp to use a nominal torque value of say 10ft-lbs and click two to three times (or whatever other value gives the most consistent value across most users), then to set the actual clamping force continue to turn the handle through an angle proportional to the desired clamping force, most easily found through experimentation as it depends on the rigidity of the vice etc etc. By clicking multiple times you get to a point where the mechanism more reliably sets the applied torque, by lubricating the vice you reduce the amount of force you're loosing to friction but also make it more consistent. By turning a fixed degree further you move the vice jaws closer together by a distance fixed by the pitch of the vice screw and by moving that fixed distance the "spring" in the clamping setup will apply a more consistent clamping force. Now, for the force applied to dislodge the part what you're measuring there is purely a function of friction. Surface area and pressure in this context are completely irrelevant, on the condition that the part will support the applied load. Basically, for two given materials with a flat contact area there will be a constant coefficient of friction and as you increase the applied force the friction will increase proportionally and independent of surface area. The reason the serrated jaws increase the friction is that by biting into the surface they change the coefficient of friction, the increase has a limit though and that's due to the shear strength of the material being clamped. At some point those little divots you're biting against in the material will deform or just shear off entirely. The mechanical failure of the surface does however depend on surface are, you either have enough friction that the material fails, or you have enough clamped surface area and material strength for the applied force that friction is overcome, or to make it more complex in practice I expect the failure of the material results in a decrease in the coefficient of friction and the part slides.
The amount of friction in the vise screw and ways will affect the test results as well. Try testing multiple vises to see how much they vary. Also the test setup needed to be very secure to support heavy loads. A part and fixture can deflect or change shape under heavy load which would affect your results. The smaller surface area of the serrated jaws allows the aluminum to compress more until it bottoms out. The super glue test it attached to a plastic substrate so it will deflect a long way before releasing. Good food for thought. Thanks for sharing.
Very good video! I think the aluminum discrepancies between smooth jaws and the grip jaws is down to 2 different factors. Firstly (Just an idea, not a fact) it could be that the aluminum is galling on your steel smooth jaws, increasing sticktion. Perhaps try different jaw materials with an aluminum workpiece? Second, I think the grip jaws could be cutting pieces out of the aluminum because its so soft. Could you please try once more the smooth versus talon/grip jaws but with different workpieces? I'm thinking aluminum, mild steel, bronze, halfhard alloy steel (4140 pht?). Also, I'm super interested if gripping length changes how the jaws grip, like if there's a big difference between the MMM jaws (full serrated) and the premade Talon Grip jaws from Mitee Bite with differing numbers of "Cleats" (basically partially serrated) Thanks for making great content btw, I look forward to your next vid
I guess you need to start testing tapes now! What tolerance does each popular tape / super glue combo add on before you factor in your actual CNC tolerance? Perhaps graph tape-type /area curves... This was awesome - so interesting. Thanks guys
Thinking about it - we should consider cutting forces that are generated with bits, feeds, speeds and material. Then you can create a zone map of force generated and really have a clear idea of what clamp approach is appropriate for what kind of cut on each material. Huge endeavour, but a selection of cut forces generated with various setups would be super interesting, I'm sure. Maybe you need to collab with high speed vid to capture the data... Unless the load cell can be connected to data recorder and synched.
@@adama1294 I know the green stuff is powder coating tape. They stated that often enough. What I want to know how well it compares to Kapton, as Kapton is cheaper.
Great video but you missed one test. When testing lateral force your block was about half the width of the vice. While stiff, even this type of vice will rack so instead of thinking of the jaws a parallel think of them as forming a trapezoid. So you're forcing a part towards the narrow side from the wide side. You could test a block the full width of the jaws so the force will be maximum at the middle, over the screw, and tapering off at the edges (limited by the rigidity of the jaws). Alternatively, you could test a 1/3 width block on each side and in the middle of the jaws. You should also monitor the stop on the left side (is that a tenths indicator?) to see how much it is deflecting or sliding. Ideally, that is your reference surface and doesn't move at all but that's not likely.
What an excellent video with exciting opportunities. I guess the biggest remaining unknown now is what is the magnitude of forces imparted to the parts during various different machining processes? This might allow you to establish typical factors of safety that you are working with and maybe allow you to intelligently estimate how hard to push things if a particular part is very difficult to hold.
I have two requests: 1) load the cell with a c-clamp and compare to your test rig. A 6'Jorgensen c-clamp puts out about 1200 lbs force/pressure (the shoe on the acme screw is arguably one square inch). I think your single point loading is skewing the results, despite the supposed rigidity of any set of jaws. 2) I'd like to see the results of using some manual-machinist vise set-ups, like using brass rod or halved ball-bearings to single-point load one side of the workpiece. Very interesting video, thank you.
Have you thought about friction when making your math? Yes, the pressure is lower with bigger surface, but neither the piece nor the jaws are totally flat... So the friction between them increase with the contact surface area. I think this is how you can explain the results you are showing today.
With the exception of certain materials (like rubber), the contact pressure is not taken into account when calculating friction force. It is only based on the normal force aka clamping force. What he is more than likely seeing is the material itself yielding. The grip jaws have those teeth in them that bite into the workpiece. That gives more resistance than just friction due to localized deformation. If he used steel as opposed to aluminum he would more than likely see the inflection point at a much higher value.
@@Worrsaint excellent points both. Bite type jaws, or inserts grip can be limited in shear or deformation of the work piece (tearing). I agree with both Thomas Manure and your self. Co-efficent of friction, for steel to steel or steel to Aluminium ranges from 0.1 to 0.3. Various tricks improve grip on milling vices, include abrasive paper, paper, copper, aluminium or similar when gripping steel, some compliance is also introduced here. Hard facing materials using Electro Spark Deposition, from Surmet, Carbonite etc, greatly improve grip.
@@bostedtap8399 From what I have seen in the comments a lot of people seem to be seeing grip and friction as one and the same. Friction is only a component of the grip. You also have compliance (as you mentioned), cold welding, van der walls forces, sticktion due to micro vacuum and just the strength of the material causing yielding in the shear direction.
@@Worrsaint I certainly had not considered Van der Walls forces, explaining the reason that smooth jaws as well as co-efficent of friction achieve grip, some synaptic transmission has increased 🤔 👍. It would be interesting to compare simple grip between steel and aluminium for the same jaws, modulus of elasticity differs greatly here, plus alloys across steels and aluminium. Regards John.
@@bostedtap8399 would also be curious to see the difference in the grip in the x and z directions. I would think it would be easier to pop it out in the z than in the x. For a piece sized like he has you would also have the most deflection inducing harmonics and chatter. Not to mention the cant of the jaws due to how they are tightened would cause them to be spread at the top more than the bottom. I do not know how much some of these matter for the vast majority of applications, but it would still be interesting to know.
Everyone reading, the clamping force applied to the vise jaw by turning the vise screw is is determined by 4 factors, Torque applied to the screw, Dia. of the screw, Pitch of the thread, and friction from the treads meshing. Torquing vises with equal torque that have different screws will produce different clamping forces.
I've used a thick two sided tape for UHMW fixtures mounted on aluminum. If both surfaces are clean, they hold really good. Just don't leave on too long or its a pain to get off.
In the aluminum tape/superglue/tape scenario It deflected a lot. If you were to apply force to so that it nears maximum deflection and then remove the force does the work piece return to the original position?
The talon jaws have those serrations on them so they hold with more than just friction. I suspect the aluminum is yielding at the higher forces which is why you see the drop. You would probably get a different inflection point when using steel.
The reading of the load cell will not be accurate if you apply a force over the whole face with the vise. They are to constraint in that way. Put a ball bearing the hole so you have only one point of contact per side.
this is correct, for most accurate and linear measurements. the whole S frame will deflect in a non linear fashion if used on the flats. there is a more suitable load cell out there. that type used is more of a tension device. the cylindrical ones are used most often for compression readings.
I was going to say I remember hearing the exact same thing in my labs at school but now that I’m looking through manuals and videos I can’t explicitly find someone saying “don’t load the whole face, just the bolt hole”. Most manufacturers ARE saying s beam load cells are for tension and compression just not making a point about how to do the the compression load. I would be curious to see if you repeat the torque wrench test a few times (same person same technique) with a point load and a few times with it on the whole face if the standard deviation would be significantly different? My gut says it would.
Which is why they simply recorded it as “load cell units”. There was no interest in actual units like Newton’s, or lbs. only comparative units as to what held more.
Annealed aluminum is super soft, and will have plastic deformation at about half of the pressure of copper. I'm thinking that the repeated clamping of the talon jaws into the test piece acted like heating the surface, annealing it.
Ok, now how does that relate to cutting forces? If it takes say 500 pounds to move the workpiece in the vise how does that compare to the normal cutting forces exerted by various machining operations? How much force are you willing to subject the machine to? Is this even an issue until you get to higher HP machines?
You might instruct your people not to touch a torque wrench anywhere but the handle, and if the handle has a pivot, to be sure it is floating between its stops. To brace yourself on the head of the wrench will cause inaccurate readings, as will bending the shaft of the wrench by pushing the pivoted handle against its stop. I know it is easier to get more torque on any wrench if you can brace yourself on the head of the wrench (I do it myself); the extreme of that is the cross-wrench where you double the torque by pulling on one arm while pushing on the other. However, you shouldn't do it when using a torque wrench. Try it both ways and see. (In fact, this is a request for such testing: Use the torque wrench both ways to the same torque setting, and see if there is a difference on the load cell.)
Tried super-glue and a simple piece of paper between saturating it with glue before putting on the work-piece? It would give a stand-of that may make it easier to separate the parts but that would essentially be nothing but super-glue with cellulose fiber reinforcement...
Typical holding failure doesn’t come from it sliding out of the vise. Failure is usually pull out of the vise due to the mechanical leverage of the helix angles on the cutting tools. I’d rig a load test to see how much force it takes to pop the part up out of the vise
To eliminate the slip in the vice you should turn the vice, so that the fixed jaw is holding the load. Another point is, that only the answer (p=F/A) of the specialist is only the first approximation. This is very well know e.g. for mechanical brakes an even car wheels. In general one should think, that the pressure is higher with less thick wheel because of p=F/a, But it is as so often the micro structure of the materials. You do not know the correct A. Therefore broader wheels happen to have more grip if the street is dry, not if the street is wet. If the street is wet, you have to have more pressure by reducing the contact area. Perhaps you know, that for racing teams, they like to use slicks with no profile on dry streets and change immediately to profiled tires, once it is raining, because then the pressure is higher because of less contact area. So it is not as easy. Another point for vice pressure is the question of friction within the vice. If friction in the mechanism is higher, you are not able to put much force on the work piece. So be aware, that friction is very important especially in turning machines. If the chuck is not maintained properly, you can have at high speeds, because of centrifugal forces less pressure on the part, and it may fly away. As physicist in safety sciences I had to examine many accidents (mostly lethal) in these cases.
Now for steel it would be interesting how much "clamping force" a magnet can give. A lot of people have said they cant hold shit, but form time to time i've done quite some work on the jobs with thin stock
If you have a 4" long piece and grip 1/2 " of it at 10 pounds, would that not equal a 4" piece gripping 1" deep at 5 pounds? Shouldn't equal psi be the same gripping regardless of how many square inches are clamped?
You forgot one key exact thing about torque wrenches, hand placement. Too far in or out will vary the actual torque reading, and double clicking also adds about 10-15% higher in torque value.
The amount of force exerted by the hand will change, the torque will not. The biggest variation is how fast you click it. To be accurate you need to go slow and sneak up on it. You could see them throwing their weight on it instead of gradually exerting more force until the end and over traveling the click.
I've also heard that hand placement makes a difference, even with clicking torque wrenches. Even though I've watched a video that "proved" it, I'm still a bit skeptical. It'd be very cool if the SMW crew were to test this if they make a followup video. Testing different speeds would be interesting too.
@@Hirudin We did testing with torque sensors and stuff at my old job. It does not make a difference where you hold it. The big difference was how quickly and suddenly you apply force due to the spring mechanism used to make it "click".
@@Worrsaint sorry but that is wrong, it entirely depends on the mechanism of the torque wrench. That particular wrench they are using has an effective length line on the handle, that's where the force should be applied, either side of that line and the torque applied will be wrong.
@@dj280z You twist the handle to set the proper spring tension for a given torque. The torque is effectively measured at the head of the torque wrench. Where you press along the lever changes the tangential force required to click it, not the amount of force 50 lbs 6inches out or 25lbs 12 inches from the center of the anvil rotation both give 25 ft-lbs of torque. If where on the lever you pressed mattered, it would not be measuring torque by definition.
We can se some of these results in a non machining setting, in the snow a narrow wheeled car will grip better than fat wide tires because the same weight is spread over a wider area which lowers the lb/sq". but in the same area the wide tires give better traction because they are pushing on more material/surface area almost a contradiction but its not.
Perhaps the best way to get consistency using a torque wrench is using torque-angle. Torque to a low number (maybe 10-20lb./ft.) and then turn a certain number of degrees. This is how many fasteners are tightened in automotive applications these days. The problem is that the more clamping force you have, the less accurate the torque wrench is; especially when stopping and starting (having to overcome starting friction). You can buy torque angle wrenches and repeat the test!
there is another part, and that is time, i.e. put the unit under pressure and then leave it for a bit, because you may have the part beginning to fail but you increase the pressor that that then makes it look like it fails at the higher value. when you are machining a part, you get is to fail just you start to machine or a short time later into the machining.
would still want to hold a good portion of the work if you were machining something very tall as the side pressures exerted at the top of the piece by the tool would act as a fulcrum at the upper most point of the jaws which is holding the work. doesn't seem like it would be too hard to torque the part out of the jaws at that point.
So basically the only way to get more than 500ish lbs of lateral force is to use superglue?? Interesting... I guess I don’t need to rush out and get those serrated jaws. I assumed they would increase the workholding power tremendously. How do the dovetail and 5th axis vises compare?
Nice video.. thanks for all your knowledge. I really need your help on cutting a AR15 bolt. I am trying to mill off some of the lower part. This thing is Soooooo hard.. what can I use?. I have a mini mill what do I need to do to cut this..please help...
If you search for "hard milling" you'll probably find some helpful videos. The key is you need to use a tool that is harder than your material. I'm curious, what do you need to mill the bolt for?
Vice bolt pitch and friction has alot to do with varying values when using the torque wrench. To say 80ftlb will give 5000 lbs clamping on every vice is misleading. No way.
For Test 4 you said you are putting the same amount of force on a smaller surface area which should result in higher pressure. You are confusing the pressure of the load cell and the pressure of the vise. You said you would expect a higher load cell to move the block but you just did this test and found that not to be true. What I think is really happening is, with a smooth jaw you are relying on friction to hold the part in place but with the talon grips you are biting into the material which is not a friction but a more mechanical stop which is why it takes more force. I am not sure why the force required is then lower at higher torques. I wish you would have done the engineering calcs before hand to have an expected value. This shouldn't be to hard. Take the force of the vise (normal force) multiply by the coefficient of friction and that would be the force required to move the block with the load cell.
@@LambertZero Yes of course, this only this only applies to somewhat flat surfaces. If we allow for deformation or interlocking shapes all bets are off.
Friction is the function of the surface roughness; material; and force. Area has no part in it which is why you were getting the same values for holding thinner pieces as thicker.
Even with the inconsistencies in the torque wrench its a lot better then handing the vise handle to someone and telling them to make it tight then they ask how tight😂
Just some quick thoughts. Firstly, coulomb friction (generally a good approximation for non moving things) is independent of the affected area ( F=Fn*u u being a constant). With those serrated jaws the part deforms and the theeth actully dig in (I can't tell if it was elastic or plastic, but I'd assume elastic). If the part was only deformed elasticly under low/ medium clamping Force, it would explain the decreasing holding power under higher forces. At some point the material yields and subsequently tension is released.
Rakh c it has both elastic and plastic deformation... I’m sure some of it springs back but there are definitely teeth marks in the part after clamping.
@@steves.3485 Would be fun to test the prestamped parts with serrated jaws. Like LANG has the stamping machine to stamp the toothmarks into (harder) materials, then vise doesn't have to be tightened so strongly to get the holding power. But I'd guess this doesn't work well with aluminium, it'll just rip out of the serrations with enough force.
In main , this is not important . Most deceive is how part will deform and re deform after unclamping . Indicator is essential to avoid goin out of narrow range of tolerance . The best way is to weld a part . That's why aluminium is not so easy to machining , has to be machin and cutoff .
“Torque wrenches aren’t accurate or repeatable”. Not when your using a $38 POS, go buy yourself a real torque wrench, I wouldn’t even buy a $38 breaker bar, nevermind torque wrench.
So what we learned here is that Alex, while looking like a guy, has the touch of a small woman. He's guy you want use when you don't want the result to be broken and deformed.
John, love the video. Long time watcher, first time commenter. Im not a machinist so you’re always way over my head But as someone who has worked in force/load cell calibration for years, you need to be mindful that the load cell needs to be loaded at the points where it is threaded. Each side of that s-cell is a lever and as it deforms by design to give the signal that the strain gauges pick up and convert to force readings the fulcrum point changes when force is applied to the flat surface and applied load changes. Using the threaded holes and a bolt through them ensures that the fulcrum point is in the same place as when it was calibrated, increasing your accuracy.
Very interesting info. Thanks for that.
This is the type of video that really makes you stop and think how many different variables a programmer has to consider when parts could be ran by multiple operators. Keep up the good work :)
Based on your setup, there is a very simple answer as to why clamping area on smooth jaws did not affect the force required to move the part (hint, it has nothing to do with pressure).
The force applied by the jaws acts 90 degrees (perpendicular) to the "pushing" force you are applying. Therefore the 2 different forces act completely independently of each other on part (think drawing the forces as X and Y directions on a graph). Then what keeps the part from immediately moving? Friction! The friction force between the jaw face and part is what keeps the part in place. If we knock the dust off our handy engineering textbook, the formula for calculating friction force is: Friction=(perpendicular force)X(coefficient of friction). Notice how the formula does not contain anything about surface area, that is because as counter intuitive as it is, friction is completely independent of surface area and there is our answer.
For fun, we can actually calculate the coefficient of friction between your vice jaws and the part using the numbers from your experiment. You clamped the part with about 3300lbs of force and it took about 550 lbs to move the block so if we plug those number in we get: 530=3300X(coefficient of friction) so the coefficient equals about 0.17!
However, this only holds true for the very specific setup you created here. You still want to hold on to as much area as you can because cutting forces on the part never act just in a singular direction like your test setup. The force most likely to dislodge your part during machining is from torque or tipping force from the tool pushing on the edge of the part. To best counter this, you'd want to reduce the lever arm from which the torque is applied by leaving the minimum amount of stock exposed.
Very well put Mech engineer. Was gonna basically post the same thing along with my post above about why the variation when using the torque wrench but luckily having seen your post that already explained it, I didn't have to take the time.
Considering the other force directions, do you think it can be defined the minimum area required for an specific part to be in contact with the smooth jaws in order to stay still and not move?
I have NEVER seen this discussed anywhere. Bravo for the experiments, insight, and numbers. Just continued awesomeness.
AVE
Someone in your shop abides! New information has come to light man.
Friction force isn't a function of the Area: F =µ*N (µ: Friction coefficient, N: Normal force). I guess Talon jaws add plastic deformation to the equation which isn't very effective at high forces as Aluminium is quite soft and therefore deforms easily
I was wondering about the same thing. Talon jaws grip really well, but aluminium just is too soft to hold on. Now you'd need to test smooth surfaces vs gripped jaws with different materials :)
John, having tested several torque wrenches a few years ago, I found that the number one cause for error is how you use it when it comes to clicker type torque wrenches. Clicker type wrenches are not good for when you want repeatable accurate torque because of the way they work. They are really more of situations when speed is of more of a concern than accuracy and and being repeatable. If one does not instantly let of pressure the moment the click happens, you will get a bit more torque applied than intended but this is really a relatively small amount of induced error in most cases.
Where the real problem with them is though is the way you pull on the handle. If you are not pulling PERFECTLY perpendicular to the axis of rotation of the fastener and within a few degrees of perpendicular to the beam of the wrench, you will get more torque before the click occurs than intended. This is because of the friction induced in the hinge that allows the beam of the wrench to slightly deflect when the detent releases. Basically the detent doesn't roll at the set spring tension because friction of the pivot is preventing the detent from seeing the real force being applied.... Essentially, friction at the pivot is acting as a path for force to bypass the detent.
If not quite following what I mean by perfectly perpendicular to the axis of rotation, think of it in the sense of this video and using the torque wrench to tighten the vise. if you are applying any force to the handle that is either towards you (the operator) or away from you (basically force that is in the same direction as the fastener, you are adding friction to the pivot which means that the detent is not seeing the real force, but rather a force that is less than what you are really applying.
For illustration purposes, have a look at this illustration.
en.wikipedia.org/wiki/Torque_wrench#/media/File:Direct_pawl_clicker_torque_concept.png
if you are applying any force that is outside of perpendicular axis of rotation of the fasten, that force is creating friction on the pivot pin which in effect is causing the assembly to act more like a rigid member. Until enough force is applied to overcome this friction, the detent (red square peice) never sees the real force. So lets say that you are generating enough friction that it takes 10lbs of force to overcome the friction, that is 10lbs more force that is being used to create torque on the fasten.. Essentially, its like you added 10lbs of force the the spring
Beam type torque wrenches are best if you want repeatable and accurate (if its a quality and accurately calibrated wrench) torque values but at the expense of speed. Beam type wrench are unaffected by forces that are outside of being perpendicular to the axis of rotation (as long as the beam needle is still floating). This is assuming that you cant use the both stretch measurement method (which is really is the most accurate method obviously).
Hope this helps for anyone to understand why / how there could be so much variation from person to person and even for two different values from the same person.
For the drop off in holding power for the griping jaws you may be hitting the yield strength of the aluminum part itself due to the side pressure acting over a smaller area
Hi John,
Have you tried or heard of "Electro Spark Deposition", you can have grip faces in steels, surface coated with a very hard textured/abrasive media, imagine gluing 40 to 100 grit abrasive cloth/paper to grip surfaces. Excellent in cutting through mill scale, and can be applied to any shape surface. Rocklanise, Surmet or Carbinite are the suppliers that come to mid.
Great subject.
I think you'd want high clamping force, but low clamping pressure. That means grip as large an area as you can afford. The higher the pressure, the more you will damage the surface.
True, the clamping force is what is most important and will be what holds the part in place, pressure is bad in this case as it will lead to dings in the part.
Very interesting, for the torque wrench, I noticed each person used it differently, very differently, also the Min-Max delta, or range was massive.
Like the data though!.
Here's an idea in case you decide to do another video: test how consistent the clutch settings on electric drills are. And/or try an impact wrench (electric or pneumatic) anf a couple of those extensions that are supposed to limit torque.
Great info in this video, it is much appreciated!
I have learned much from your videos. Thank you. In a prior life I owned a company that built high performance engines, etc. for mostly Porsche's. I just wanted to make some observations.
1. I always use the same method to tighten bolts using a torque wrench. That is to first get the fastener barely tight so that the pull for the actual torque setting can be done in a single smooth stroke; stopping the click or beep. Also, the pull direction should be in an arc that is on the same plane as the rotation of the screw/bolt, ie., do not jiggle it forward or rearward.
2. Manufacturers almost always provide torque specs on the high side. 225 ft lbs on a pull stud is huge. I have seen the number used in some threads. I have a HAAS TL-1 lathe and HAAS TM-3P mill. The mill has a spindle holdfast included. If I put even close to 225 ft lbs on it, I am sure it would rip the unit off the machine. Not to mention that using a 1/2" torque wrench to do this would be outside af many folks strength. Similar to asking any but the most experienced machinists to measure anything at the .0001" level even using the proper mic.
3. My experience tells me that applying too much torque to a part is a much more prevalent and damaging problem than not enough. Using a 6" x 4" x 1/2" piece of 6061 T6 aluminum gripped 1/4" deep along the 6" side using smooth jaws, in a ORANGE vise, the workpiece will begin to deflect at about 45 ft lbs (on a calibrated torque wrench). At 100 ft lbs, the deflection at the centerline is more than .002". The worst part is that if you face cut that part, you would not even notice. If you measure all 4 corners you would;d not notice. If you measure the center line, you will notice.
4. You hit the nail on the head. Even with proper use and a quality torque wrench, you will get different results. If you include the torque wrenches that most people buy, results become inconsistent (the worst) and wrong. I am not sure if most folks return their torque wrench setting to the lowest setting after each use.
5. I know there are countless machinists out there that know what to torque their work pieces. Too bad that I have never seen a post anywhere like this:
"I have a high quality torque wrench that was recently calibrated, an ORANGE smooth jaw vise, HAAS Mill and run at high speeds and feeds. Here are some example torque values I currently use:"
Then, they could list several work pieces that they mill and include the vise torque values, metal type, size, gripping type and area, cutter info, speeds and feeds, etc. IF I HAD THESE REAL WORLD EXAMPLES, I COULD EASILY FIGURE THE REST OUT.
One thing about holding on a small amount of material, is that you will end up with deformation on your part, also if your part is up out of the vise more your tool will have a easier time rotating the part out rather than pushing it’s all about leverage.
Interesting about the torque wrench variations. Good video for my apprentice tomorrow. Thanks for doing the experiment!
This is really great work you're doing and sharing with the rest of us and I truly appreciate it, thank you. But I haven't noticed a lot of issues with parts moving sideways in the vise. It's typically pulled up. I imagine if you do the same series of tests but with a jack screw to apply upward force you'll see very different results more aligned with what we would expect. ( drill a clearance hole for the jack screw through the fixture to test super glue).
Next you'll have to test what kind of forces are put on the part during cuts, climb vs conventional, slot milling vs side milling, rigid tapping, and of course how that changes with tool size. Keep up the great work looking into the science of machining.
Please try an ice plate and a vacuum chuck.
Thanks, JIM
The thing I found most impressive about using superglue to fixture a part is the tapes holding strength. Makes me want to see the differences between different types of tape.
Excellent and informative video. Being old school T&D guy. We used traditional vise handle with a pipe...if we needed heavy gripping. Wow were we wrong....
Did it work out though? If it worked, I'd say you're right too.
Hey John, check this out. I used to work for a Top Fuel Drag Racing Team that was sponsored at various times, by Snap-On, Matco, Cornwell Tools, and Travers Tool Company. There was a toolbox in the machine shop that had probavly 75 torque wrenches in it, high dollar Snap-On, Matco, etc. There was also a digital torque torque tester mounted on the wall by the box. Out of all of the torque wrenches, I believe only a total of 2 were accurate. Accepted practice was to set the wrench to your value needed, say 80 ft. lbs., and keep adjusting the torque wrench, until you hit the number required. So, even though the Snap-On (most expensive) cost a buttload of money, it was no more accurate than anything else.
What else may be overlooked is the actual mating surface geometry ( geometric tolerances) The modulus of elasticity of the vise is fairly constant, since the materials are similar. The part being held , aluminum , is less , and hence will distort elastically against the smooth mating face of the hardened steel jaw much sooner. The vise is just an irregularly shaped spring, and will deform as the applied load increases. Therefore, there are a lot of unseen forces that will alter the shape of the held part that just aren’t perceivable. You essentially are machining a squished part ( elastic deformation) , and the forced aren’t just in compression, but an increase in buckling as the contact area is less and less as the part is gripped higher in the vise. So, as already stated, the lowest pressure on the faces of the part is ideal, with a margin of safety.
1) keep part from moving or releasing during cut.
2) keep part from being too elastically , hopefully not plastically deformed while held during machining.
Just plain great!
Why didn't you rotate the vise 90 degrees for the superglue tests, to eliminate the vise holding onto the backing as a factor?
Thank you John. Test 2. Come up with some correlation for appropriate clamping force before a part will bow or distort. Subject to part thickness or cross-sectional area. I've always wanted to understand this better.
John....good start to understand how much work holding force you have using different methods and inputs. I would suggest when you try a new variable/process, do it more than once. This will help you see trends that you don't see with a sample size of 1. With enough measurements, you will start see the standard deviation of the process which will tell you how much variability (force) there is at confidence level. Or simply stated, do you have a robust work holding method.
I would suggest looking at your work holding stiffness rather than ultimate clamping force. You don't necessarily care how much force the vise puts on the part as long as it holds it in place. However if are trying to do precise machining, you would want a stiff work holding method so it doesn't add to your tolerances.
Could you do similar tests with Mitee bite fixture clamps? It would be great to know a range of RDOCs and ADOCs with speeds and feeds to go with the different clamp sizes.
Great Video. But It would be more practical to know how much clamping force is needed? How hard to tighten your vice when using a 10mm endmill adaptive? Does calculating cutting force is enough or is vibration and dynamic load loosen up the part at this force?
On the torque wrench: buy a quality unit, they're not that expensive to get a good one. Make sure you pull perpendicularly to the handle from the hand grip and try not to induce twist. Come up on the desired setting very smoothly and slowly, and trust your instrument, no need to double click. Besides, if you do it right and try for a second click, it should not move the bolt at all as static friction is higher than the dynamic friction. Edit: I was further watching the video and saw on a few occasions where things were tightened up with somewhat outstretched arms resulting in the spanner jerking as it was tightened. In those cases use a longer wrench and hold it closer to your body if possible to try and get absolute smoothness when tightening
Cool experiment!
So, first thing is that the amount of clamping load your vice jaws impart is a function of several things. The screw pitch, applied torque but also, and very importantly, the friction present within the vice assembly.
If you want to substantially increase the repeatability of the clamping load the jaws impart then you need to do two things, first is to lubricate the vice screw and sliding surfaces.
Second is to change the procedure you're using to clamp to use a nominal torque value of say 10ft-lbs and click two to three times (or whatever other value gives the most consistent value across most users), then to set the actual clamping force continue to turn the handle through an angle proportional to the desired clamping force, most easily found through experimentation as it depends on the rigidity of the vice etc etc.
By clicking multiple times you get to a point where the mechanism more reliably sets the applied torque, by lubricating the vice you reduce the amount of force you're loosing to friction but also make it more consistent. By turning a fixed degree further you move the vice jaws closer together by a distance fixed by the pitch of the vice screw and by moving that fixed distance the "spring" in the clamping setup will apply a more consistent clamping force.
Now, for the force applied to dislodge the part what you're measuring there is purely a function of friction. Surface area and pressure in this context are completely irrelevant, on the condition that the part will support the applied load.
Basically, for two given materials with a flat contact area there will be a constant coefficient of friction and as you increase the applied force the friction will increase proportionally and independent of surface area.
The reason the serrated jaws increase the friction is that by biting into the surface they change the coefficient of friction, the increase has a limit though and that's due to the shear strength of the material being clamped. At some point those little divots you're biting against in the material will deform or just shear off entirely. The mechanical failure of the surface does however depend on surface are, you either have enough friction that the material fails, or you have enough clamped surface area and material strength for the applied force that friction is overcome, or to make it more complex in practice I expect the failure of the material results in a decrease in the coefficient of friction and the part slides.
The amount of friction in the vise screw and ways will affect the test results as well. Try testing multiple vises to see how much they vary. Also the test setup needed to be very secure to support heavy loads. A part and fixture can deflect or change shape under heavy load which would affect your results. The smaller surface area of the serrated jaws allows the aluminum to compress more until it bottoms out. The super glue test it attached to a plastic substrate so it will deflect a long way before releasing. Good food for thought. Thanks for sharing.
Very good video! I think the aluminum discrepancies between smooth jaws and the grip jaws is down to 2 different factors. Firstly (Just an idea, not a fact) it could be that the aluminum is galling on your steel smooth jaws, increasing sticktion. Perhaps try different jaw materials with an aluminum workpiece? Second, I think the grip jaws could be cutting pieces out of the aluminum because its so soft. Could you please try once more the smooth versus talon/grip jaws but with different workpieces? I'm thinking aluminum, mild steel, bronze, halfhard alloy steel (4140 pht?). Also, I'm super interested if gripping length changes how the jaws grip, like if there's a big difference between the MMM jaws (full serrated) and the premade Talon Grip jaws from Mitee Bite with differing numbers of "Cleats" (basically partially serrated)
Thanks for making great content btw, I look forward to your next vid
id like to see a comparison of the clear path servos and o drive
I guess you need to start testing tapes now! What tolerance does each popular tape / super glue combo add on before you factor in your actual CNC tolerance? Perhaps graph tape-type /area curves...
This was awesome - so interesting. Thanks guys
Thinking about it - we should consider cutting forces that are generated with bits, feeds, speeds and material.
Then you can create a zone map of force generated and really have a clear idea of what clamp approach is appropriate for what kind of cut on each material.
Huge endeavour, but a selection of cut forces generated with various setups would be super interesting, I'm sure. Maybe you need to collab with high speed vid to capture the data... Unless the load cell can be connected to data recorder and synched.
How does the powder coating tape compare to kapton tape? At least where I'm it's a lot cheaper.
That green tape is the same as power coating tape. It is both polyester. Kapton is the orangeish brown color.
@@adama1294 I know the green stuff is powder coating tape. They stated that often enough. What I want to know how well it compares to Kapton, as Kapton is cheaper.
I would like to see you measure the movement of the solid and moveable jaws under clamp loads. And see if you find it to be repeatable.
Great video but you missed one test. When testing lateral force your block was about half the width of the vice. While stiff, even this type of vice will rack so instead of thinking of the jaws a parallel think of them as forming a trapezoid. So you're forcing a part towards the narrow side from the wide side. You could test a block the full width of the jaws so the force will be maximum at the middle, over the screw, and tapering off at the edges (limited by the rigidity of the jaws). Alternatively, you could test a 1/3 width block on each side and in the middle of the jaws. You should also monitor the stop on the left side (is that a tenths indicator?) to see how much it is deflecting or sliding. Ideally, that is your reference surface and doesn't move at all but that's not likely.
What an excellent video with exciting opportunities. I guess the biggest remaining unknown now is what is the magnitude of forces imparted to the parts during various different machining processes? This might allow you to establish typical factors of safety that you are working with and maybe allow you to intelligently estimate how hard to push things if a particular part is very difficult to hold.
I have two requests: 1) load the cell with a c-clamp and compare to your test rig. A 6'Jorgensen c-clamp puts out about 1200 lbs force/pressure (the shoe on the acme screw is arguably one square inch). I think your single point loading is skewing the results, despite the supposed rigidity of any set of jaws. 2) I'd like to see the results of using some manual-machinist vise set-ups, like using brass rod or halved ball-bearings to single-point load one side of the workpiece. Very interesting video, thank you.
There is no difference between manual machining clamping scenarios. A brass road or screwy ball is used in any type of machining, CNC or manual.
Have you thought about friction when making your math? Yes, the pressure is lower with bigger surface, but neither the piece nor the jaws are totally flat... So the friction between them increase with the contact surface area. I think this is how you can explain the results you are showing today.
With the exception of certain materials (like rubber), the contact pressure is not taken into account when calculating friction force. It is only based on the normal force aka clamping force. What he is more than likely seeing is the material itself yielding. The grip jaws have those teeth in them that bite into the workpiece. That gives more resistance than just friction due to localized deformation. If he used steel as opposed to aluminum he would more than likely see the inflection point at a much higher value.
@@Worrsaint excellent points both.
Bite type jaws, or inserts grip can be limited in shear or deformation of the work piece (tearing). I agree with both Thomas Manure and your self. Co-efficent of friction, for steel to steel or steel to Aluminium ranges from 0.1 to 0.3. Various tricks improve grip on milling vices, include abrasive paper, paper, copper, aluminium or similar when gripping steel, some compliance is also introduced here. Hard facing materials using Electro Spark Deposition, from Surmet, Carbonite etc, greatly improve grip.
@@bostedtap8399 From what I have seen in the comments a lot of people seem to be seeing grip and friction as one and the same. Friction is only a component of the grip. You also have compliance (as you mentioned), cold welding, van der walls forces, sticktion due to micro vacuum and just the strength of the material causing yielding in the shear direction.
@@Worrsaint I certainly had not considered Van der Walls forces, explaining the reason that smooth jaws as well as co-efficent of friction achieve grip, some synaptic transmission has increased 🤔 👍.
It would be interesting to compare simple grip between steel and aluminium for the same jaws, modulus of elasticity differs greatly here, plus alloys across steels and aluminium.
Regards John.
@@bostedtap8399 would also be curious to see the difference in the grip in the x and z directions. I would think it would be easier to pop it out in the z than in the x. For a piece sized like he has you would also have the most deflection inducing harmonics and chatter. Not to mention the cant of the jaws due to how they are tightened would cause them to be spread at the top more than the bottom. I do not know how much some of these matter for the vast majority of applications, but it would still be interesting to know.
Everyone reading, the clamping force applied to the vise jaw by turning the vise screw is is determined by 4 factors, Torque applied to the screw, Dia. of the screw, Pitch of the thread, and friction from the treads meshing. Torquing vises with equal torque that have different screws will produce different clamping forces.
Nice video John. Thanks
what about different vices and lead screws?
I've used a thick two sided tape for UHMW fixtures mounted on aluminum. If both surfaces are clean, they hold really good. Just don't leave on too long or its a pain to get off.
Alex should read the manual on the torque wrench - especially the part about where to hold it!
In the aluminum tape/superglue/tape scenario It deflected a lot. If you were to apply force to so that it nears maximum deflection and then remove the force does the work piece return to the original position?
I think the aluminum started to deform and that is the deflection that you are seeing on the serrated jaws.
The talon jaws have those serrations on them so they hold with more than just friction. I suspect the aluminum is yielding at the higher forces which is why you see the drop. You would probably get a different inflection point when using steel.
The reading of the load cell will not be accurate if you apply a force over the whole face with the vise. They are to constraint in that way. Put a ball bearing the hole so you have only one point of contact per side.
this is correct, for most accurate and linear measurements. the whole S frame will deflect in a non linear fashion if used on the flats. there is a more suitable load cell out there. that type used is more of a tension device. the cylindrical ones are used most often for compression readings.
@@nilzlima3027 also very correct
I was going to say I remember hearing the exact same thing in my labs at school but now that I’m looking through manuals and videos I can’t explicitly find someone saying “don’t load the whole face, just the bolt hole”. Most manufacturers ARE saying s beam load cells are for tension and compression just not making a point about how to do the the compression load.
I would be curious to see if you repeat the torque wrench test a few times (same person same technique) with a point load and a few times with it on the whole face if the standard deviation would be significantly different? My gut says it would.
Splitting hairs here. We're talking a few % difference at most.
Which is why they simply recorded it as “load cell units”.
There was no interest in actual units like Newton’s, or lbs. only comparative units as to what held more.
Hi. Vill we meet on EMO Hannover this week ?
Nice thank you.
Lance & Patrick.
Annealed aluminum is super soft, and will have plastic deformation at about half of the pressure of copper. I'm thinking that the repeated clamping of the talon jaws into the test piece acted like heating the surface, annealing it.
Very interesting, John..
Ok, now how does that relate to cutting forces? If it takes say 500 pounds to move the workpiece in the vise how does that compare to the normal cutting forces exerted by various machining operations? How much force are you willing to subject the machine to? Is this even an issue until you get to higher HP machines?
How about testing a vacuum chuck? Don't you have a Pierson plate?
You might instruct your people not to touch a torque wrench anywhere but the handle, and if the handle has a pivot, to be sure it is floating between its stops. To brace yourself on the head of the wrench will cause inaccurate readings, as will bending the shaft of the wrench by pushing the pivoted handle against its stop. I know it is easier to get more torque on any wrench if you can brace yourself on the head of the wrench (I do it myself); the extreme of that is the cross-wrench where you double the torque by pulling on one arm while pushing on the other. However, you shouldn't do it when using a torque wrench. Try it both ways and see. (In fact, this is a request for such testing: Use the torque wrench both ways to the same torque setting, and see if there is a difference on the load cell.)
Tried super-glue and a simple piece of paper between saturating it with glue before putting on the work-piece? It would give a stand-of that may make it easier to separate the parts but that would essentially be nothing but super-glue with cellulose fiber reinforcement...
Typical holding failure doesn’t come from it sliding out of the vise. Failure is usually pull out of the vise due to the mechanical leverage of the helix angles on the cutting tools.
I’d rig a load test to see how much force it takes to pop the part up out of the vise
So how did you release the final part that was directly superglued, without damaging a theoretical part?? Great vid...
Heat breaks down glue.
Go check it clickspring, he uses lots of superglue "arbors" for working with brass.
To eliminate the slip in the vice you should turn the vice, so that the fixed jaw is holding the load. Another point is, that only the answer (p=F/A) of the specialist is only the first approximation. This is very well know e.g. for mechanical brakes an even car wheels. In general one should think, that the pressure is higher with less thick wheel because of p=F/a, But it is as so often the micro structure of the materials. You do not know the correct A. Therefore broader wheels happen to have more grip if the street is dry, not if the street is wet. If the street is wet, you have to have more pressure by reducing the contact area. Perhaps you know, that for racing teams, they like to use slicks with no profile on dry streets and change immediately to profiled tires, once it is raining, because then the pressure is higher because of less contact area. So it is not as easy.
Another point for vice pressure is the question of friction within the vice. If friction in the mechanism is higher, you are not able to put much force on the work piece. So be aware, that friction is very important especially in turning machines. If the chuck is not maintained properly, you can have at high speeds, because of centrifugal forces less pressure on the part, and it may fly away. As physicist in safety sciences I had to examine many accidents (mostly lethal) in these cases.
Now for steel it would be interesting how much "clamping force" a magnet can give. A lot of people have said they cant hold shit, but form time to time i've done quite some work on the jobs with thin stock
If you have a 4" long piece and grip 1/2 " of it at 10 pounds, would that not equal a 4" piece gripping 1" deep at 5 pounds? Shouldn't equal psi be the same gripping regardless of how many square inches are clamped?
On the direct superglue, heat separates, acetone cleans the glue off.
Love you sir, love your content.....♥️♥️♥️💫
You forgot one key exact thing about torque wrenches, hand placement.
Too far in or out will vary the actual torque reading, and double clicking also adds about 10-15% higher in torque value.
The amount of force exerted by the hand will change, the torque will not. The biggest variation is how fast you click it. To be accurate you need to go slow and sneak up on it. You could see them throwing their weight on it instead of gradually exerting more force until the end and over traveling the click.
I've also heard that hand placement makes a difference, even with clicking torque wrenches. Even though I've watched a video that "proved" it, I'm still a bit skeptical. It'd be very cool if the SMW crew were to test this if they make a followup video.
Testing different speeds would be interesting too.
@@Hirudin We did testing with torque sensors and stuff at my old job. It does not make a difference where you hold it. The big difference was how quickly and suddenly you apply force due to the spring mechanism used to make it "click".
@@Worrsaint sorry but that is wrong, it entirely depends on the mechanism of the torque wrench. That particular wrench they are using has an effective length line on the handle, that's where the force should be applied, either side of that line and the torque applied will be wrong.
@@dj280z You twist the handle to set the proper spring tension for a given torque. The torque is effectively measured at the head of the torque wrench. Where you press along the lever changes the tangential force required to click it, not the amount of force 50 lbs 6inches out or 25lbs 12 inches from the center of the anvil rotation both give 25 ft-lbs of torque. If where on the lever you pressed mattered, it would not be measuring torque by definition.
We can se some of these results in a non machining setting, in the snow a narrow wheeled car will grip better than fat wide tires because the same weight is spread over a wider area which lowers the lb/sq". but in the same area the wide tires give better traction because they are pushing on more material/surface area almost a contradiction but its not.
try a kurt vise maybe?
Hey John are you at EMO 2019 ?
Perhaps the best way to get consistency using a torque wrench is using torque-angle. Torque to a low number (maybe 10-20lb./ft.) and then turn a certain number of degrees. This is how many fasteners are tightened in automotive applications these days. The problem is that the more clamping force you have, the less accurate the torque wrench is; especially when stopping and starting (having to overcome starting friction). You can buy torque angle wrenches and repeat the test!
Is it vise in USA or a vice
Vise is the holder and vice is associated with police. Unless you use the word like vice chairman, vice president then it is the next person down.
there is another part, and that is time, i.e. put the unit under pressure and then leave it for a bit, because you may have the part beginning to fail but you increase the pressor that that then makes it look like it fails at the higher value. when you are machining a part, you get is to fail just you start to machine or a short time later into the machining.
would still want to hold a good portion of the work if you were machining something very tall as the side pressures exerted at the top of the piece by the tool would act as a fulcrum at the upper most point of the jaws which is holding the work. doesn't seem like it would be too hard to torque the part out of the jaws at that point.
So basically the only way to get more than 500ish lbs of lateral force is to use superglue?? Interesting... I guess I don’t need to rush out and get those serrated jaws. I assumed they would increase the workholding power tremendously.
How do the dovetail and 5th axis vises compare?
Спасибо ! Очень наглядно и познавательно
Nice video.. thanks for all your knowledge. I really need your help on cutting a AR15 bolt. I am trying to mill off some of the lower part. This thing is Soooooo hard.. what can I use?. I have a mini mill what do I need to do to cut this..please help...
If you search for "hard milling" you'll probably find some helpful videos. The key is you need to use a tool that is harder than your material.
I'm curious, what do you need to mill the bolt for?
Vice bolt pitch and friction has alot to do with varying values when using the torque wrench. To say 80ftlb will give 5000 lbs clamping on every vice is misleading. No way.
Enjoyed...
For Test 4 you said you are putting the same amount of force on a smaller surface area which should result in higher pressure. You are confusing the pressure of the load cell and the pressure of the vise. You said you would expect a higher load cell to move the block but you just did this test and found that not to be true. What I think is really happening is, with a smooth jaw you are relying on friction to hold the part in place but with the talon grips you are biting into the material which is not a friction but a more mechanical stop which is why it takes more force. I am not sure why the force required is then lower at higher torques. I wish you would have done the engineering calcs before hand to have an expected value. This shouldn't be to hard. Take the force of the vise (normal force) multiply by the coefficient of friction and that would be the force required to move the block with the load cell.
Static friction is dependent on the friction coefficient of the materials and the normal force, not surface area.
In vacuum yes. In real life, however, serrated jaws bite in and deform the workpiece, effectively changing the coefficient of friction.
@@LambertZero Yes of course, this only this only applies to somewhat flat surfaces. If we allow for deformation or interlocking shapes all bets are off.
Friction is the function of the surface roughness; material; and force. Area has no part in it which is why you were getting the same values for holding thinner pieces as thicker.
Yes, finally someone points out this basic relationship .
Even with the inconsistencies in the torque wrench its a lot better then handing the vise handle to someone and telling them to make it tight then they ask how tight😂
Interesting that someone would test whether a longer lever provides more advantage....
It's interesting you think that is what you saw.
@@Hirudin It's what he said and what I saw
This is one of the Videos that makes me love the Internet...
I love Ed’s new blonde hair. 👍
Just some quick thoughts. Firstly, coulomb friction (generally a good approximation for non moving things) is independent of the affected area ( F=Fn*u u being a constant). With those serrated jaws the part deforms and the theeth actully dig in (I can't tell if it was elastic or plastic, but I'd assume elastic). If the part was only deformed elasticly under low/ medium clamping Force, it would explain the decreasing holding power under higher forces. At some point the material yields and subsequently tension is released.
Rakh c it has both elastic and plastic deformation... I’m sure some of it springs back but there are definitely teeth marks in the part after clamping.
@@steves.3485 Would be fun to test the prestamped parts with serrated jaws. Like LANG has the stamping machine to stamp the toothmarks into (harder) materials, then vise doesn't have to be tightened so strongly to get the holding power. But I'd guess this doesn't work well with aluminium, it'll just rip out of the serrations with enough force.
reflection on vices fixed jaw by clamping pressure
In main , this is not important . Most deceive is how part will deform and re deform after unclamping . Indicator is essential to avoid goin out of narrow range of tolerance . The best way is to weld a part . That's why aluminium is not so easy to machining , has to be machin and cutoff .
Julie needs to eat more steak ! 😉
You have a sample size of 1 for each person. So you can't really take anything away from the results.
You should teach your guy how to use a tork key. Hey giving 20000lbs after the "clic"
l.o.l
One of your younger guys could take this farther as a thesis project.
u spying on fireball tools or what ;)
“Torque wrenches aren’t accurate or repeatable”. Not when your using a $38 POS, go buy yourself a real torque wrench, I wouldn’t even buy a $38 breaker bar, nevermind torque wrench.
*these* data...
So what we learned here is that Alex, while looking like a guy, has the touch of a small woman. He's guy you want use when you don't want the result to be broken and deformed.
Zero. Because they have force, not power. Butchering these poor units.
It's not units. Units would be like units of length, inches, feet, meters, etc. Units of power and units of force, but force and power aren't units.
It’s vise not vice.
Harry, adjusted and thx oeps on my side (weak excuse ... Dutch is my native language)
Nice advise.
@@pesterenan pun intended?
It’s oops, not oeps. 😉
Torque wrenches are supposed to be plus or minus 5% or 10%, depending on who makes it. You're looking at a 30+% difference.