John, just a suggestion. (I don't know on the TorMach machine having never run one.) On most all CNC machines if you program you finish cuts using cutter comp you can control the amount of stock you are leaving or removing by changing the offset for the diameter or radius (some machines use diameter some use radius) in the tool offset page for that tool. For instance in your cam software if you don't offset the tool path to centerline tool path but apply a comp left or right (G41 if traveling on the left side G42 on the right side, you post should automatically put the code in the program). You will need lead in and out moves equal to at least the radius of the intended tool size (I usually add a little more to the radius say .010"). Than you can than just rerun the finish cuts at the machine adjusting the offset to come to size. The way you appear to be doing it you have to go back to the computer to reprocess code every time(what if the tool wears or you change to a new tool). That does work but lets say you have many machines with many operators it could be very difficult to do if one person is programming. Or if you run a job many years later you would have to remember what to do. Or using reground end mills. So it is better to program the actual shape of the part and use the tool offsets to come to size than every one will know how to run each program without needing any additional programing or processing of code. Also looking at a program latter on in the G code it is much easer to tell what size you are cutting.The Haas machine will do this.
When you do your 2d contour to clean up, in the passes tab, change the Compensation type from 'In Computer' to 'In Control'. Also don't leave any stock. Add +.002 to the diameter offset of the tool doing the contour in your CNC control. Run the program, measure the slot, then change the wear offset to what is left to cut and run the program again. You don't have to repost the program and that tool is now dialed in. Then you can also easily adjust this as the tool wears out.
Hey John, when you run a CAM simulation, switch the material from 'mirror' to 'wall paint'. It makes things MUCH easier to see. Thanks for another great video. Ken
I've seen a few generic case hardened answers, however, carburization is actually something that's very interesting. You end up heating the low carbon steel while it's wrapped in an air tight 'chamber' and surrounded by charcoal or other high carbon elements. If you heat long enough, you can actually get then almost completely through hardened. It's a very ancient technique that hasn't changed much over time, and essentially very very effective.
Dude! You need to wring your gauge blocks together. ...and, the gauge block carnage as they fly off onto the VM table and rattle around! The horror! Oh! ...the horror!!!
1018 CRS can be case hardened by carberizeing. It usually yields a hardened depth of .060 to .065 of an inch.at 56 to 58 Rd. Not very many places do it any more though. I made parallels that way and some were so hard a carbide end mill wouldn't cut them. Keep up the good work John
1018 can be case hardened. I made a planer gauge in school and had it case hardened to 55-60 Rockwell. My understanding is that carbon is added to the material from the outside before heat treat.
If I remember correctly, you are correct. Multiple kings had their highest guests of honor used aluminum silverware rather than gold. Aluminum itself isn't really toxic, but they didn't have any way to process it back then, so it would have other contaminants with it. Currently aluminum is more expensive than steel weight wise, however steel is harder and better for more structural builds. I knew a robotics team that milled out their entire chassis out of aluminum and spent close to $15,000, however their robot was extremely light and almost 30 pounds under the weight limit with more moving parts than most teams. Another team had a chassis out of steel, but they had to "swiss cheese" their robot to fit within the weight limit, but they didn't have as much functionality. The steel chassis cost them almost $600, but very cheap compared to the aluminum version.
if you use a super quench (water, salt and surfactants) from a little over 1550 degrees you can get it to the 42-45 range. It works by basically breaking up the steam bubble that forms around it to increase contact area of the quench liquid to the part. The 50 range is probably not happening on 1018 unless there is a method I am not aware of.
Great video John, made my own set of Talon grip jaws today, did not hardened them thou, works great. If you make jaws 1.875" wide, you can make other workholding fixture on other side.
as others have mentioned, you can case harden 1018. Most likely it was qpq'd (salt bath nitirding). It will usually have a black finish after that process.
Use 20MnCr5 for the talons. And case harden them to 55-60HRC , about 0.5mm case depth should be just fine.. They will bite into the workpiece. But that is why these talons work i guess. It's the angle on talons that matter. If you need any other suggestions, just ping. We run a machine shop and one of our other divisions is a heat treatment division.
I haven't seen those jaws before so don't know what their hardness claims involve, but I case harden parts of 1018 or 1020. I've never bothered measuring the hardness but can if you would like me to? But they are definitely more than "file hard" as I always check that after hardening. The advantage of case hardening is the soft core, yet very hard case. It's also cheaper steel ... a biggie when you're as cheap as I am!
58c is obtainable through carborizing. We induction harden 1018 to mid 50's and can achieve deep effective case hardness. The majority of the parts we hear treat are 1020 or 4140 though.
Nice work John. It's funny, I was having the same identical issue just last night trying to center a feature using the fx measurements. I wrote it off as not being able to use the derived measurement belonging to another component (in my case), but now I'm not so sure since you were working with just one component. I hope you will post an update for the rest of us if you find out what the deal was. Ciao, Marco.
You need to water quench or brine quench. 1018 will be ok. But if you control the process properly , you just might get 55 hrc. Better take 20MnCr5. Assured you will get 55Hrc+
dare I say: Gripping video? P.S. Slide the gage blocks together for better repeatability. If you just stack them, you trap air. Good test is to see if the thin block falls off under its own weight. It should be held there by air having trouble getting between the blocks.
Bernd Felsche I was thinking the same about the gauge blocks. If you are looking for accurate measurements, the blocks should slide together and stick like you said. :)
NYC CNC Surfaces of the blocks are so flat that they contact each other so close they form an atomic covalent bond as well as a Van Der Waals bond (what makes geckos able to stick to anything) . Kinda like magnetism except they're held together by sharing electrons instead of actual magnetic forces. This is also why if you leave two blocks wrung together just sitting around, they'll form a near solid block as if welded together. You may actually break the blocks before breaking the surface bond when trying to get them apart. So never leave blocks wrung together. Including ceramic and carbide blocks. Both will form covalent bonding between them.
If 1018 is the alloy it must be case hardened. The 18 in 1018 is the nominal carbon content. It will be around 0.18%, but needs double or triple to hold a higher hardness.
Geof Dumas look up super quench. while 1018 is generally considered to not be hardenable, with the right surfactants and stuff in the quench it is possible to harden into the low 40s. It usually makes sense to just go for a higher carbon steel, but it is actually possible to do.
Make a clamp wrench that you fix to each vice 'permanently' so it's right there on each vice all the time, no searching and no pulling off one to put on the other. I know you plan to have three or four clamps in there so have a tightening bar on each.
hey I have an idea for a new series. it would be an xcarve series. you buy an xcarve assemble it and then put it through its paces and then upgrade it until it becomes a usable small scale production machine. (then of course you can sell the upgrades)
There was a hinky bit there in the middle. How many times did you have to run that 0.499" pass to get it to come out on size? 'Cause it looked like you were having a hell of a time with the gage blocks there in the middle, no narration, just stack after stack of gage blocks. You can case harden low carbon steels either through carburizing or by using a hardening compound (Kasenit, Cherry Red, etc).
You're trying to do a driven dimension you can do it yet you can do this (d8-d7)/2 this works just did it. Yes that worked fine for the sketch, But why did you not just sketch the jaw from the side, then draw a centerline on the top then a line for the front converted to construction
Any way that works is correct, but just for the record... I would have used pins or expanding parallels and a micrometer to "project" those two inner planes into an external place where I could get them with a micrometer, that only works at the ends of the slot tho, cant measure the middle as easily.
They used to think that aluminium pots, anti-perspirants etc. contributed to Alzheimer's disease but more modern medical research has disproven that link. However patients undergoing kidney dialysis should not use aluminium cooking equipment as the dialysis process causes the aluminium to build up inside the body where it can take the place of calcium in the body, potentially leading to bone disease and some forms of anemia.
Don't think they can get 1018 to 50+Rc. Why no surfacing pass so your slot isn't crooked with your top? I mean sure it is just .002 off but still that seems a lot on a precision vise surface.
bcbloc02 If you go to the talon jaws site they say they are only within.002. Talons are really only for 1st op only so .002 out of parallel doesn't really matter as it's holding onto saw cut or roughly stock anyway.
Great vid John. I don't entirely trust my 1100's accuracy and have to sneak up on things where the dimensions really matter which is a bit of a pain. If I use the same tool, same fixture etc I can easily shave off 0.01 mm (1/2 thou) accurately by running the cycle again with negative stock to leave so its repeatability seems really good. However I can't trust it to interpolate a hole say 15.01 mm reliably - it'll be off 0.02 or more one way or the other. So I guess it is precise rather than accurate. It'd be interesting to see how your Haas and Tormach compared in terms of precision and accuracy - the Haas at my trade school is supposedly accurate (or maybe precise) to 0.001mm (half a tenth).
What I do the keep from making multiple programs to "sneak up" on a dimension is lie to the machine and tell it the diameter of the tool is touch larger than it really is, this way the cutter radius comp cuts a tad smaller then you check fit and adjust tool diameter as needed. Of course all this assumes you are using cutter radius compensation. Further more once you get it close, throw it on the surface grinder and dial it in.
Charles Reiche I'll have to double check. But I believe if you use the "create hole" feature to put a hole in an existing solid, there's a pull down menu that lists all the number and letter drill sizes, along with tap drill sizes for corresponding thread sizes.
Hi John. When you adjust the last thou on your Talon Jaw slots, are you going back to the PC and change the values or have you cutter compensation G41/ G42 direct in the cnc mill (compensation type: in control) and only change the cutter diameter in the tool table of your machine?
Miteebite has the solid right on their website just need to make an account. They have it for most of their products. Super handy when making dedicated fixturing. I've seen you use fusion 360 a few times now and I can't say I prefer the software. Although it is definitely a cheaper solution to mastercam and solidworks.
Being a Physics and Mathematics degree type with Engineering on top - Pls consider a binary search on the test blocks. e.g. - .4 to .5 easy .4 not .5. Cut in half. .45 works ? no .425 fits - try 4.3 and see if it is to much. Al can be absorbed into the body and can cause brain issues. It is a 'light' metal so it isn't as Cd or Pb (lead). Remember the run away from cooking dishes in raw Al ?
I'm at my parents house while the wife is sleeping at home in prep for her night shift. feeding the baby while watching this. flex arm starts tapping- dad in the other room "hey that sounds like a flex arm!!! I know that sound a mile away" #HeHasTappedAFewHoles
Why does the drill start pecking so far above the material? Doesn't the machine know exactly where the tip of the bit is? Or is it just a safety margin, so if it's a bit off, it won't jam into the stock?
It converts to a number cause you selected a reference dimension. A reference dimension is there just for that, reference. A reference number should never be used to calculate derived dimensions.
Correct me if I am wrong but as you measured with the Vernieer, why chase the measurement with the gage blocks why not check with the talons you were going to put in the slot? (sorry i wrote the comment whilst watching!)
Tony Perks Digital/Dial/Vernier calipers are commonly referred to as "Very Near" calipers especially on internal measurements. Used correctly gage blocks or if applicable, adjustable parallels and a micrometer is a much more precise method of measuring slot width and being able to see the incremental increase in size between cuts. This will allow precise cut adjustments to finish the slot to the desired dimension.
Cad-Cam_Man Maybe he means instead of just using the gauge blocks why not use the talons that are going to fit in the slot to check the slot as well as using the huge blocks. (Which h did do, but he made the comment before watching the full video.)
NYC CNC I can, but I've submitted several bugs, several months ago, none of which have gone anywhere. I still think fusion has incredible value, I just look forward to the day it's a little more stable.
@NYC CNC Apparently it's not possible to use a driven dimension as a parameter just yet, for ...reasons... forums.autodesk.com/t5/design-validate-document/use-driven-dimension-as-a-parameter/td-p/6253015 My guess (as a professional software engineer) is that it'll be something to do with the possibility of introducing circularly referential dimensions. That's just a guess though.
The carnage of them flying out of his hands, rattling around off of the table and vise, and down into the chip pan, too! Not once, ...but twice!! I bet the head of QC (if they have one) just cringed, if and when they watched this video.
At the 20:51 spot on the video I see on the side of the slot you cut 8 regularly spaced darker spots. Does this suggest something wonky with the drive in that axis?
I think you are right. I thought he was going for 6 holes but I was mistaken and the "blemish" doesn't appear until after he drills. Good catch. Thanks!
I have been looking into getting a set of Talon Jaws for my tow Kurt DX6 vices. I believe that they work good but are they worth it? I do not do large production runs, mostly prototyping with a few runs when we get the model right. Right now I have to use a lot wider stock to hold it in the vice so the part can be cut out basically doing pocket milling. What I am not sure about is how hard is it to machine off the bottom .060" left form holding it in the jaws? I can make several of the same parts using a form of nesting in my CAM software which is a SolidWorks plug in from MecSoft called VisualCAM for SolidWorks. So I have a longer piece of material to cut the remaining stock from. Also I am not sure if some of the parts would be able to be held correctly to face off the remaining stock. Anybody have thoughts/ideas on this? Cost of the jaws are not a problem. I am just wondering if they are worth it.
How come when you use the Haimer to Zero the workpiece, the indicator rotates 2x to zero. Don’t you just touch the workpiece and as soon as it moves you’ll know its the edge of the workpiece?
When it touches the workpiece its -2mm (or something equivilent int the imperial version), it becomes a lot more clear if you watch the bottom dial! :)
The jaws I know are soft 1018 steel. I didn't think the actual clamps were 1018 though. Max hardness for 1018 is around 42-45 Rc Unless maybe they are hard carbon coating them to 52Rc and calling that surface hardened. But no. 1018 can't get that hard by itself, it's carbon content is too low to break the low to mid 40's on hardness Your "D8 -D7" won't work because D8 is a driven dimension that was projected. You have to use the D# that was used in the first sketch to create the initial jaw. At least in my experience algorithmic dimensioning doesn't play nice with driven dimensions You Really really need to wring your gauge blocks when measuring. Yes it's a PITA to do every time, but unwrung blocks could measure big by a thou or two
I have to ask the question. As a machinist with your capabilities where do you draw the line on intellectual property and copyright infringement? As a business owner you can machine a great deal of items just like your vises that are for sale? I understand your intent here and we all hate paying the buck sometimes I enjoy your shop talks and your business model shows not as much as your machining shows but they hold a value to me so how do you prevent from being sued by talon? As a machinist in your apartment in New York and as a Business owner have those lines changed?
Joshua McClain As long as you are not directly profiting from those items (i.e. Selling them) there isn't any infringement. Just like carve smart and Lang both give out the prints on how to manufacture their soft and hard jaws, even though they are patented. Again. As long as you are not selling them directly it's basically fair use. Otherwise you'd have to pay a royalties fee to manufacture and sell that design
That's the problem Jon is selling them...while not the talon clamp yet..his "machinist clamps" he is selling a product he reverse engineered...at least this time he credited who's design he ripped off....not much more ethical however....what's next a saunders machine Technigrip??? I find I'm quickly losing respect for this channel.
If you are referring to the 'kant twist' clamps, it depends on the copyright, Patent date & time since. How many machinists with talent don't make many of their own tools... Got to be a realist. He is not putting anyone out of business.
Doug Bourdo Kant twist design patent has long since expired The name is however trademarked so he can't sell them as "Kant twist" clamps. Only as cantilever clamps.
I was referring to the ethics question of if you should do it, not if you can legally get away with it....yes you can get away with this..but should you, and what will people think of your practices.
Good parts and fine job, but it's proof positive to be that for onesie-twosies manual machines are WAY better then CNC. - Hell, we could have even finish GRIND the slot to a few tenths. CNC has it's place but this ain't it. - Could actually have made this part without actually having a drawing at all let alone having to spend time on a 3D model.
There is very little evidence of toxicity of aluminium or using aluminium cutlery and dishes and pots. It is possible to have bad effects, but you need significant amount consumed as salts i.e. in water. As a metal or the amount that will be in the food will be thousands times smaller than risky dose, even if you scratch the oxide layer with metal tools. I always was worried about this too, but in reality it is not an issue. The aluminium just feels cheap these days, and not too heavy, so people don't have too much confidence about it, and maybe reminds them of lead in how it looks. Not sure.
why would anyone buy a pallet for 99$? I mean, anyone who has a milling machine to use it can just drill and tap those 80 holes pretty quick. If they were made from steel and hardened i could understand, as not everyone can harden and surface grind their own plate. All the best tho!
meocats Same reason people buy soft jaws from monster jaws... Work out the cost to manufacturer a few one off pallets in house vs buying for $99. I guarantee you the in house cost is close to double the buy cost.
4:45-4:50. You mention using the outside of the "jaw" versus using the traditional depth gauge on the opposite side. How do you do that and get an accurate measurement?
This is starting to bother me not necessarily the lack of hass videos but the thought that you have to have a machine running as fast as possible if you are going to post a video of it. We get it if you are given the "perfect" tooling and a whole bunch of time you can make a machine go really fast. I would have liked to see you post a video, how ever slow it may be, of the first time you ran that part then explain how much time it take to tune and push everything to its limits, and then post a video of how fast you make a part. This is one the main things that bothers me, I get why y'all do it you want to make yourself and the machine look good or so I think. But when you post a video of a machine running in the perfect environment you make machining look easy when it actually takes mouths to get your cycle times that low. No disrespect just one of my rants, I really do like your videos Mason
daniel lyall I don't mean that. What I am talking about is when he said I don't want to post the hass videos till he gets it to where he is pushing the tools and the machine. Don't remember which video he said that in though.
I don't know the theory behind any "adaptive" strategies, but am I the only one that thinks there is something wrong? The tool spends at least 50% of its time cutting air. How is this efficient or better in any way? I know slotting is hard on tools taking a 100% diameter cut, but one pass down the middle and one to each side and the slot is done, probably in 1/2 the time this took (not factoring in sneaking up on the width.) In John Grimsmos last video he even commented on trying to reduce tool path wasted time by eliminating as much unnecessary movement. If there is a good reason for all that cutting air then let me know. I am definitely no expert in this at all. It just seems like a "gee whiz isn't that neato" button on fusion 360. It's "adaptive", it must be better.... That is a lot of wasted horsepower, especially on the Haas. Also, why not rapid on the air cutting part?
From what little I understand it reduces heat and therefor increases tool life by controlling the amount of material that is in contact with the tool at any one time. I imagine it also reduces tool deflection as well. I suppose the opposing idea is that in a saturated work environment where every second extra that a CNC is being used is costing time that it could be used on another project and thus tool life isn't an issue rather cutting speed is.
It's easier on the tool, but you are not wrong about "cutting air" being wasted time. Fusion doesn't seem to have any way to improve the linking moves to minimize this wasted time (or maybe John just doesn't know how to use them, and since I don't use Fusion yet, neither do I). I know a lot of other CAM packages rapid through linking moves and do other things to minimize wasted time. The biggest "problem" in John's videos that make these adaptive strategies really inefficient is the fact that Tormach machines have quite slow rapids (something like 125-150 IPM). So even if he's got it outputting rapids for the linking moves, he's still capped at 150 IPM, which is crawling along when you're cutting air. Strictly speaking, on a real VMC that's not even a fast cutting feed rate. When your rapids are 1,000 or 1,500 IPM, those linking moves don't take so much time, so the toolpath is a lot more productive (you'll get more like 85% time in-cut VS the Tormach's ~50% time in-cut for the same tool path). My VMC is ancient, with rapids of only 300 IPM, so when I program, I try mightily to keep the tool down and in the cut as much as possible. That means deep cuts with wide step-overs and slower feed rates. It's old school, but it IS more productive on a machine with slow rapids. These adaptive toolpaths really don't start to break even productivity wise until you've got rapids in the 700-800 IPM range. Any slower than that, and all the time spent out of the cut on linking moves kills your productivity, and the savings in tool wear doesn't even come close to covering the lost metal cutting. You'll really see the adaptive start to shine on his new HAAS though, if he ever gets around to doing something interesting with it while the cameras are rolling.
On a faster VMC or any other machine for that matter, you won't have as much "air cutting" time since the rapid feeds will be much higher than the tormac can handle.
Barry Gerbracht It could/should be faster. The limitation is more the machine having only a max feed rate of 100ipm and not the best accel/deccel rates. Most adaptive paths will allow climb and conventional cutting rather than climb only. But again, machine HP and rigidity prevents being able to take the heavier faster cuts. So he's not really "wasting" any HP just doesn't have much to start with. And yes, full width slotting in a high HP machine with heavier castings and drive motors could rough that slot in 2 passes faster than an adaptive path. But for the tormach the path he used here is the most optimal.
Barry Gerbracht The theory behind the adaptive toolpath is something called trochoidal milling. essentially the idea is that the toolpath controls radial cutter engagement keeping it constant. The benefit there being that you can push axial engagement so that cuts that would take multiple depths can be done in one, and increase feed rates and RPMs based on radial chip thinning. Ideally the air cutting passes should be done at the maximum linear feed rate of the machine, unless the machine has accel/deccel problems. Air cutting should not be done in rapid feed since rapid disables both linear and circular interpolation, and depending on the machine may not be as precise.
John, just a suggestion. (I don't know on the TorMach machine having never run one.) On most all CNC machines if you program you finish cuts using cutter comp you can control the amount of stock you are leaving or removing by changing the offset for the diameter or radius (some machines use diameter some use radius) in the tool offset page for that tool. For instance in your cam software if you don't offset the tool path to centerline tool path but apply a comp left or right (G41 if traveling on the left side G42 on the right side, you post should automatically put the code in the program). You will need lead in and out moves equal to at least the radius of the intended tool size (I usually add a little more to the radius say .010"). Than you can than just rerun the finish cuts at the machine adjusting the offset to come to size. The way you appear to be doing it you have to go back to the computer to reprocess code every time(what if the tool wears or you change to a new tool). That does work but lets say you have many machines with many operators it could be very difficult to do if one person is programming. Or if you run a job many years later you would have to remember what to do. Or using reground end mills. So it is better to program the actual shape of the part and use the tool offsets to come to size than every one will know how to run each program without needing any additional programing or processing of code. Also looking at a program latter on in the G code it is much easer to tell what size you are cutting.The Haas machine will do this.
True. He should've used another endmill for "finishing". Just needed that fresh edge.
When you do your 2d contour to clean up, in the passes tab, change the Compensation type from 'In Computer' to 'In Control'. Also don't leave any stock. Add +.002 to the diameter offset of the tool doing the contour in your CNC control. Run the program, measure the slot, then change the wear offset to what is left to cut and run the program again. You don't have to repost the program and that tool is now dialed in. Then you can also easily adjust this as the tool wears out.
Hey John, when you run a CAM simulation, switch the material from 'mirror' to 'wall paint'. It makes things MUCH easier to see. Thanks for another great video. Ken
Awesome. I was very glad to learn you can hold the alt key to pick a single part of a contour! I will use this daily. Thanks!
I've seen a few generic case hardened answers, however, carburization is actually something that's very interesting. You end up heating the low carbon steel while it's wrapped in an air tight 'chamber' and surrounded by charcoal or other high carbon elements. If you heat long enough, you can actually get then almost completely through hardened. It's a very ancient technique that hasn't changed much over time, and essentially very very effective.
It is super awesome that your 1100 can generate such precise results after all these years
This is a perfect situation for a use of cutter compensation. Nice video 👍👍👍
my old boss invented the talon jaws. I miss machining. love your videos
Dude! You need to wring your gauge blocks together. ...and, the gauge block carnage as they fly off onto the VM table and rattle around! The horror! Oh! ...the horror!!!
1018 CRS can be case hardened by carberizeing. It usually yields a hardened depth of .060 to .065 of an inch.at 56 to 58 Rd. Not very many places do it any more though. I made parallels that way and some were so hard a carbide end mill wouldn't cut them. Keep up the good work John
I love how the whole machine moves while the table stays very stable 😁
probably cause the camera is mounted on the table which is moving...
So.. I know this video is old, but I just have to say, AWESOME job using freggin hot keys. I hate all the extra clicks
1018 can be case hardened. I made a planer gauge in school and had it case hardened to 55-60 Rockwell. My understanding is that carbon is added to the material from the outside before heat treat.
If I remember correctly, you are correct. Multiple kings had their highest guests of honor used aluminum silverware rather than gold. Aluminum itself isn't really toxic, but they didn't have any way to process it back then, so it would have other contaminants with it. Currently aluminum is more expensive than steel weight wise, however steel is harder and better for more structural builds. I knew a robotics team that milled out their entire chassis out of aluminum and spent close to $15,000, however their robot was extremely light and almost 30 pounds under the weight limit with more moving parts than most teams. Another team had a chassis out of steel, but they had to "swiss cheese" their robot to fit within the weight limit, but they didn't have as much functionality. The steel chassis cost them almost $600, but very cheap compared to the aluminum version.
Aluminum was also used for the cap on the Washington monument, about that same time frame.
if you use a super quench (water, salt and surfactants) from a little over 1550 degrees you can get it to the 42-45 range. It works by basically breaking up the steam bubble that forms around it to increase contact area of the quench liquid to the part. The 50 range is probably not happening on 1018 unless there is a method I am not aware of.
Great video John, made my own set of Talon grip jaws today, did not hardened them thou, works great. If you make jaws 1.875" wide, you can make other workholding fixture on other side.
You can case harden 1018 to those hardness, but the hardened area only goes .010"-.050" deep depending on their process.
as others have mentioned, you can case harden 1018. Most likely it was qpq'd (salt bath nitirding). It will usually have a black finish after that process.
Use 20MnCr5 for the talons. And case harden them to 55-60HRC , about 0.5mm case depth should be just fine.. They will bite into the workpiece. But that is why these talons work i guess. It's the angle on talons that matter. If you need any other suggestions, just ping. We run a machine shop and one of our other divisions is a heat treatment division.
I haven't seen those jaws before so don't know what their hardness claims involve, but I case harden parts of 1018 or 1020. I've never bothered measuring the hardness but can if you would like me to? But they are definitely more than "file hard" as I always check that after hardening. The advantage of case hardening is the soft core, yet very hard case. It's also cheaper steel ... a biggie when you're as cheap as I am!
58c is obtainable through carborizing. We induction harden 1018 to mid 50's and can achieve deep effective case hardness. The majority of the parts we hear treat are 1020 or 4140 though.
Nice work John. It's funny, I was having the same identical issue just last night trying to center a feature using the fx measurements. I wrote it off as not being able to use the derived measurement belonging to another component (in my case), but now I'm not so sure since you were working with just one component. I hope you will post an update for the rest of us if you find out what the deal was. Ciao, Marco.
you can carborize low carbon steel and reach 60 before the draw down. (case harden)
You need to water quench or brine quench. 1018 will be ok. But if you control the process properly , you just might get 55 hrc. Better take 20MnCr5. Assured you will get 55Hrc+
dare I say: Gripping video?
P.S. Slide the gage blocks together for better repeatability. If you just stack them, you trap air. Good test is to see if the thin block falls off under its own weight. It should be held there by air having trouble getting between the blocks.
Bernd Felsche I was thinking the same about the gauge blocks. If you are looking for accurate measurements, the blocks should slide together and stick like you said. :)
NYC CNC
Surfaces of the blocks are so flat that they contact each other so close they form an atomic covalent bond as well as a Van Der Waals bond (what makes geckos able to stick to anything) . Kinda like magnetism except they're held together by sharing electrons instead of actual magnetic forces.
This is also why if you leave two blocks wrung together just sitting around, they'll form a near solid block as if welded together. You may actually break the blocks before breaking the surface bond when trying to get them apart.
So never leave blocks wrung together. Including ceramic and carbide blocks. Both will form covalent bonding between them.
Pretty cool. Btw, the Vizsla pup will be here this Friday. :D
Adjustable parallels would be nice to get a good measurement and after that use gauge blocks to check for taper.
If 1018 is the alloy it must be case hardened. The 18 in 1018 is the nominal carbon content. It will be around 0.18%, but needs double or triple to hold a higher hardness.
Agreed.
Geof Dumas look up super quench. while 1018 is generally considered to not be hardenable, with the right surfactants and stuff in the quench it is possible to harden into the low 40s. It usually makes sense to just go for a higher carbon steel, but it is actually possible to do.
Low 40's maybe, but not 50's, as listed
Make a clamp wrench that you fix to each vice 'permanently' so it's right there on each vice all the time, no searching and no pulling off one to put on the other. I know you plan to have three or four clamps in there so have a tightening bar on each.
1018 can be case hardened
hey I have an idea for a new series. it would be an xcarve series. you buy an xcarve assemble it and then put it through its paces and then upgrade it until it becomes a usable small scale production machine. (then of course you can sell the upgrades)
Enjoyed the Vid John!
There was a hinky bit there in the middle.
How many times did you have to run that 0.499" pass to get it to come out on size?
'Cause it looked like you were having a hell of a time with the gage blocks there in the middle, no narration, just stack after stack of gage blocks.
You can case harden low carbon steels either through carburizing or by using a hardening compound (Kasenit, Cherry Red, etc).
You're trying to do a driven dimension you can do it yet you can do this (d8-d7)/2 this works just did it.
Yes that worked fine for the sketch, But why did you not just sketch the jaw from the side, then draw a centerline on the top then a line for the front converted to construction
Any way that works is correct, but just for the record... I would have used pins or expanding parallels and a micrometer to "project" those two inner planes into an external place where I could get them with a micrometer, that only works at the ends of the slot tho, cant measure the middle as easily.
They used to think that aluminium pots, anti-perspirants etc. contributed to Alzheimer's disease but more modern medical research has disproven that link.
However patients undergoing kidney dialysis should not use aluminium cooking equipment as the dialysis process causes the aluminium to build up inside the body where it can take the place of calcium in the body, potentially leading to bone disease and some forms of anemia.
John, so after all these years, what is your go-to vice jaw ? still talons ? thanks Joe
Don't think they can get 1018 to 50+Rc. Why no surfacing pass so your slot isn't crooked with your top? I mean sure it is just .002 off but still that seems a lot on a precision vise surface.
bcbloc02
If you go to the talon jaws site they say they are only within.002.
Talons are really only for 1st op only so .002 out of parallel doesn't really matter as it's holding onto saw cut or roughly stock anyway.
Great vid John. I don't entirely trust my 1100's accuracy and have to sneak up on things where the dimensions really matter which is a bit of a pain.
If I use the same tool, same fixture etc I can easily shave off 0.01 mm (1/2 thou) accurately by running the cycle again with negative stock to leave so its repeatability seems really good. However I can't trust it to interpolate a hole say 15.01 mm reliably - it'll be off 0.02 or more one way or the other. So I guess it is precise rather than accurate.
It'd be interesting to see how your Haas and Tormach compared in terms of precision and accuracy - the Haas at my trade school is supposedly accurate (or maybe precise) to 0.001mm (half a tenth).
I wonder if you have some backlash on your Y ballscrew? Could explain why the dimension was so far under initially.
Tell me a little about current status of that Aurdino CNC you made a little while ago...!!
1:37 No, you can't harden 1018 steel. 57 Rockwell doesn't describe A, B, or C scale. Maybe this is A or B and that's just how hard it is anyway?
Of course Judd wants to go outside!
What I do the keep from making multiple programs to "sneak up" on a dimension is lie to the machine and tell it the diameter of the tool is touch larger than it really is, this way the cutter radius comp cuts a tad smaller then you check fit and adjust tool diameter as needed. Of course all this assumes you are using cutter radius compensation. Further more once you get it close, throw it on the surface grinder and dial it in.
They should totally make it so you could dimension a circle with a Number drill diameter, like 'D#21'.
Charles Reiche
I'll have to double check. But I believe if you use the "create hole" feature to put a hole in an existing solid, there's a pull down menu that lists all the number and letter drill sizes, along with tap drill sizes for corresponding thread sizes.
you ever think about getting some sliding parallels, ill sell ya a set,even a planer shaper guage for big keyways and slots
I'm so jealous of this guy
Hi John. When you adjust the last thou on your Talon Jaw slots, are you going back to the PC and change the values or have you cutter compensation G41/ G42 direct in the cnc mill (compensation type: in control) and only change the cutter diameter in the tool table of your machine?
Miteebite has the solid right on their website just need to make an account. They have it for most of their products. Super handy when making dedicated fixturing.
I've seen you use fusion 360 a few times now and I can't say I prefer the software. Although it is definitely a cheaper solution to mastercam and solidworks.
You can case harden 1018 and a36 pretty hard
The 1018 was definitely case-hardened
Man if only Fusion 360 had a hole wizard
those are awesome!
But how am I gonna get all those sweet TH-cam upvotes by telling you that 2/3rds of your jobs should start in Talon Jaws? =)
Being a Physics and Mathematics degree type with Engineering on top - Pls consider a binary search on the test blocks. e.g. - .4 to .5 easy .4 not .5. Cut in half. .45 works ? no .425 fits - try 4.3 and see if it is to much. Al can be absorbed into the body and can cause brain issues. It is a 'light' metal so it isn't as Cd or Pb (lead). Remember the run away from cooking dishes in raw Al ?
I'm at my parents house while the wife is sleeping at home in prep for her night shift. feeding the baby while watching this.
flex arm starts tapping-
dad in the other room "hey that sounds like a flex arm!!! I know that sound a mile away"
#HeHasTappedAFewHoles
Nice vid man, tnx for sharing it bro!!!
Why does the drill start pecking so far above the material? Doesn't the machine know exactly where the tip of the bit is? Or is it just a safety margin, so if it's a bit off, it won't jam into the stock?
NYC CNC oh, ok. I wasn't sure. thanks for getting back to me.
It converts to a number cause you selected a reference dimension. A reference dimension is there just for that, reference. A reference number should never be used to calculate derived dimensions.
The hardening is most likely a case-hardening.
Correct me if I am wrong but as you measured with the Vernieer, why chase the measurement with the gage blocks why not check with the talons you were going to put in the slot? (sorry i wrote the comment whilst watching!)
Tony Perks because you need a physical measured number to know how much you need to adjust the cutter by. Otherwise you could oversize the slot.
Tony Perks Digital/Dial/Vernier calipers are commonly referred to as "Very Near" calipers especially on internal measurements. Used correctly gage blocks or if applicable, adjustable parallels and a micrometer is a much more precise method of measuring slot width and being able to see the incremental increase in size between cuts. This will allow precise cut adjustments to finish the slot to the desired dimension.
Cad-Cam_Man Maybe he means instead of just using the gauge blocks why not use the talons that are going to fit in the slot to check the slot as well as using the huge blocks. (Which h did do, but he made the comment before watching the full video.)
How much is the backlash on your y-axis ballscrew?
Hi! What kind of camera do you have on your desk? (for 15 minutes)
I've noticed using the referenced dimensions in Fusion intermittently works. Another small bug that makes things very frustrating.
NYC CNC I can, but I've submitted several bugs, several months ago, none of which have gone anywhere. I still think fusion has incredible value, I just look forward to the day it's a little more stable.
Why did you switch from solidworks to fusion? Was it a cost related thing?
Thank you for your reply. I can understand that. This is the same reason when I was in university they switched from nx to solidworks many moons ago.
8:36 but do you want one on center though?
@NYC CNC Apparently it's not possible to use a driven dimension as a parameter just yet, for ...reasons... forums.autodesk.com/t5/design-validate-document/use-driven-dimension-as-a-parameter/td-p/6253015
My guess (as a professional software engineer) is that it'll be something to do with the possibility of introducing circularly referential dimensions. That's just a guess though.
Is there a specific reason for not wringing your gauge blocks together?
The carnage of them flying out of his hands, rattling around off of the table and vise, and down into the chip pan, too! Not once, ...but twice!! I bet the head of QC (if they have one) just cringed, if and when they watched this video.
Scratch and dent sale: "Been kicking around the mill for only a few weeks - almost as new" ;)
Nice!
At the 20:51 spot on the video I see on the side of the slot you cut 8 regularly spaced darker spots. Does this suggest something wonky with the drive in that axis?
ExtantFrodo2 I believe you are seeing the reflection of the holes on the side of the slot.
I think you are right. I thought he was going for 6 holes but I was mistaken and the "blemish" doesn't appear until after he drills. Good catch. Thanks!
Why didn't you tap the threads in the CNC?
I have been looking into getting a set of Talon Jaws for my tow Kurt DX6 vices. I believe that they work good but are they worth it? I do not do large production runs, mostly prototyping with a few runs when we get the model right. Right now I have to use a lot wider stock to hold it in the vice so the part can be cut out basically doing pocket milling. What I am not sure about is how hard is it to machine off the bottom .060" left form holding it in the jaws? I can make several of the same parts using a form of nesting in my CAM software which is a SolidWorks plug in from MecSoft called VisualCAM for SolidWorks. So I have a longer piece of material to cut the remaining stock from. Also I am not sure if some of the parts would be able to be held correctly to face off the remaining stock.
Anybody have thoughts/ideas on this? Cost of the jaws are not a problem. I am just wondering if they are worth it.
How come when you use the Haimer to Zero the workpiece, the indicator rotates 2x to zero. Don’t you just touch the workpiece and as soon as it moves you’ll know its the edge of the workpiece?
When it touches the workpiece its -2mm (or something equivilent int the imperial version), it becomes a lot more clear if you watch the bottom dial! :)
16:44 why didn't you make just two straight through passes
The jaws I know are soft 1018 steel. I didn't think the actual clamps were 1018 though. Max hardness for 1018 is around 42-45 Rc
Unless maybe they are hard carbon coating them to 52Rc and calling that surface hardened.
But no. 1018 can't get that hard by itself, it's carbon content is too low to break the low to mid 40's on hardness
Your "D8 -D7" won't work because D8 is a driven dimension that was projected. You have to use the D# that was used in the first sketch to create the initial jaw. At least in my experience algorithmic dimensioning doesn't play nice with driven dimensions
You Really really need to wring your gauge blocks when measuring. Yes it's a PITA to do every time, but unwrung blocks could measure big by a thou or two
The Talon Grips are A2, the jaws are 1018.
If a .501 fits in there, it's .501. Also, you need to wring the gages together to remove the clearance between them, you can't just stack them.
According to Steel Forge, 1018 can be case hardened to Rc 58.
www.steelforge.com/aisi-1018/
Thanks,
John
I have to ask the question. As a machinist with your capabilities where do you draw the line on intellectual property and copyright infringement? As a business owner you can machine a great deal of items just like your vises that are for sale? I understand your intent here and we all hate paying the buck sometimes I enjoy your shop talks and your business model shows not as much as your machining shows but they hold a value to me so how do you prevent from being sued by talon? As a machinist in your apartment in New York and as a Business owner have those lines changed?
Joshua McClain
As long as you are not directly profiting from those items (i.e. Selling them) there isn't any infringement.
Just like carve smart and Lang both give out the prints on how to manufacture their soft and hard jaws, even though they are patented.
Again. As long as you are not selling them directly it's basically fair use. Otherwise you'd have to pay a royalties fee to manufacture and sell that design
That's the problem Jon is selling them...while not the talon clamp yet..his "machinist clamps" he is selling a product he reverse engineered...at least this time he credited who's design he ripped off....not much more ethical however....what's next a saunders machine Technigrip??? I find I'm quickly losing respect for this channel.
If you are referring to the 'kant twist' clamps, it depends on the copyright, Patent date & time since. How many machinists with talent don't make many of their own tools... Got to be a realist. He is not putting anyone out of business.
Doug Bourdo
Kant twist design patent has long since expired
The name is however trademarked so he can't sell them as "Kant twist" clamps. Only as cantilever clamps.
I was referring to the ethics question of if you should do it, not if you can legally get away with it....yes you can get away with this..but should you, and what will people think of your practices.
Good parts and fine job, but it's proof positive to be that for onesie-twosies manual machines are WAY better then CNC. - Hell, we could have even finish GRIND the slot to a few tenths.
CNC has it's place but this ain't it. - Could actually have made this part without actually having a drawing at all let alone having to spend time on a 3D model.
Why not tap in the machine?
There is very little evidence of toxicity of aluminium or using aluminium cutlery and dishes and pots. It is possible to have bad effects, but you need significant amount consumed as salts i.e. in water. As a metal or the amount that will be in the food will be thousands times smaller than risky dose, even if you scratch the oxide layer with metal tools. I always was worried about this too, but in reality it is not an issue. The aluminium just feels cheap these days, and not too heavy, so people don't have too much confidence about it, and maybe reminds them of lead in how it looks. Not sure.
I don't think you a ringing the slips together correctly
why would anyone buy a pallet for 99$? I mean, anyone who has a milling machine to use it can just drill and tap those 80 holes pretty quick. If they were made from steel and hardened i could understand, as not everyone can harden and surface grind their own plate. All the best tho!
meocats
Same reason people buy soft jaws from monster jaws...
Work out the cost to manufacturer a few one off pallets in house vs buying for $99. I guarantee you the in house cost is close to double the buy cost.
Case hardened
It is RPM and not RPMs. Look it up. Look at your software and see that they don't show RPMs but instead show RPM.
Robert Evans
The number of people who actually care is you, and maybe like one other person in this world.
How did you use the edge of the caliper to measure depth?
4:45-4:50. You mention using the outside of the "jaw" versus using the traditional depth gauge on the opposite side. How do you do that and get an accurate measurement?
Slot check with an adjustable parallel?
"Adaptive control" looks like 50% air cut to me.
Otherwise, cool video.
What is the name of this tool? dl.dropboxusercontent.com/u/19129520/airtool.JPG
This is starting to bother me not necessarily the lack of hass videos but the thought that you have to have a machine running as fast as possible if you are going to post a video of it. We get it if you are given the "perfect" tooling and a whole bunch of time you can make a machine go really fast. I would have liked to see you post a video, how ever slow it may be, of the first time you ran that part then explain how much time it take to tune and push everything to its limits, and then post a video of how fast you make a part. This is one the main things that bothers me, I get why y'all do it you want to make yourself and the machine look good or so I think. But when you post a video of a machine running in the perfect environment you make machining look easy when it actually takes mouths to get your cycle times that low.
No disrespect just one of my rants, I really do like your videos
Mason
If you mean the drilling on the Haas John said about that 2 weeks ago
daniel lyall I don't mean that. What I am talking about is when he said I don't want to post the hass videos till he gets it to where he is pushing the tools and the machine.
Don't remember which video he said that in though.
Whats more irritating is people who cant spell haas.
Makes them look like a dumbhass.
I don't know the theory behind any "adaptive" strategies, but am I the only one that thinks there is something wrong? The tool spends at least 50% of its time cutting air. How is this efficient or better in any way? I know slotting is hard on tools taking a 100% diameter cut, but one pass down the middle and one to each side and the slot is done, probably in 1/2 the time this took (not factoring in sneaking up on the width.) In John Grimsmos last video he even commented on trying to reduce tool path wasted time by eliminating as much unnecessary movement.
If there is a good reason for all that cutting air then let me know. I am definitely no expert in this at all. It just seems like a "gee whiz isn't that neato" button on fusion 360. It's "adaptive", it must be better.... That is a lot of wasted horsepower, especially on the Haas. Also, why not rapid on the air cutting part?
From what little I understand it reduces heat and therefor increases tool life by controlling the amount of material that is in contact with the tool at any one time. I imagine it also reduces tool deflection as well. I suppose the opposing idea is that in a saturated work environment where every second extra that a CNC is being used is costing time that it could be used on another project and thus tool life isn't an issue rather cutting speed is.
It's easier on the tool, but you are not wrong about "cutting air" being wasted time.
Fusion doesn't seem to have any way to improve the linking moves to minimize this wasted time (or maybe John just doesn't know how to use them, and since I don't use Fusion yet, neither do I). I know a lot of other CAM packages rapid through linking moves and do other things to minimize wasted time.
The biggest "problem" in John's videos that make these adaptive strategies really inefficient is the fact that Tormach machines have quite slow rapids (something like 125-150 IPM). So even if he's got it outputting rapids for the linking moves, he's still capped at 150 IPM, which is crawling along when you're cutting air.
Strictly speaking, on a real VMC that's not even a fast cutting feed rate.
When your rapids are 1,000 or 1,500 IPM, those linking moves don't take so much time, so the toolpath is a lot more productive (you'll get more like 85% time in-cut VS the Tormach's ~50% time in-cut for the same tool path).
My VMC is ancient, with rapids of only 300 IPM, so when I program, I try mightily to keep the tool down and in the cut as much as possible. That means deep cuts with wide step-overs and slower feed rates.
It's old school, but it IS more productive on a machine with slow rapids.
These adaptive toolpaths really don't start to break even productivity wise until you've got rapids in the 700-800 IPM range. Any slower than that, and all the time spent out of the cut on linking moves kills your productivity, and the savings in tool wear doesn't even come close to covering the lost metal cutting.
You'll really see the adaptive start to shine on his new HAAS though, if he ever gets around to doing something interesting with it while the cameras are rolling.
On a faster VMC or any other machine for that matter, you won't have as much "air cutting" time since the rapid feeds will be much higher than the tormac can handle.
Barry Gerbracht
It could/should be faster. The limitation is more the machine having only a max feed rate of 100ipm and not the best accel/deccel rates.
Most adaptive paths will allow climb and conventional cutting rather than climb only. But again, machine HP and rigidity prevents being able to take the heavier faster cuts. So he's not really "wasting" any HP just doesn't have much to start with.
And yes, full width slotting in a high HP machine with heavier castings and drive motors could rough that slot in 2 passes faster than an adaptive path. But for the tormach the path he used here is the most optimal.
Barry Gerbracht The theory behind the adaptive toolpath is something called trochoidal milling. essentially the idea is that the toolpath controls radial cutter engagement keeping it constant. The benefit there being that you can push axial engagement so that cuts that would take multiple depths can be done in one, and increase feed rates and RPMs based on radial chip thinning. Ideally the air cutting passes should be done at the maximum linear feed rate of the machine, unless the machine has accel/deccel problems. Air cutting should not be done in rapid feed since rapid disables both linear and circular interpolation, and depending on the machine may not be as precise.