Hints on making Straight Edges out of Steel
Vložit
- čas přidán 20. 07. 2024
- This video contains a selection of hints on making custom, precision straight edges out of carbon steel. I created this video because of recurring viewer questions about my 1.2 meters or 4 foot precision straight edge, which is featured in the video. It was almost exclusively made with basic hand tools. It is not made from cast iron as classic straight edges are, but rather it is a welded design out of three pieces of carbon steel.
- Věda a technologie
Great content! You were recommended by Stefan!
Thank you!
Likewise :)
From time to time, as a machine installer in the manufacturing industry, I have had the pleasure of working with engineers that not only understand their profession thoroughly, but have the complete capability of constructing their projects. They not only understand what the machine has to do, but they understand the restrictions and complications that can arise in machining, welding and construction. These are the folks a person can learn from. Ive a strong feeling you may fall into this category.
Mr. Cullen, thank you very much for your kind comment. I'm not sure if I can live up to such expectations, as you know in engineering and fabricating life failure and bad solutions are much more easy to achieve than success. They say, once you've entered a profession, you think you know everything. A few years later you know that you know not too much. And with a lifetime of insights you know that you know nothing. Be that as it may, :D, thank you again.
Hi Alex, thank you for your great insight. I have been overconfident for quite sometimes until i read your comment
I am another person who came here on Stefan's recommendation, glad to have found your channel! I look forward to catching up on all your previous video's and seeing the future ones. Best Regards Sarah
Many thanks, Sarah, glad to have at least one female among my viewers :)
Sent by Stefan, good stuff!
ATB, Robin
As I said with your other comment, I am very overwhelmed, thank you, Sir. I hope not to disappoint you with my approach on this quick and dirty straight edge.
Quick and dirty, heaven help us when he really tries.
First, I must say that this is a great technique. I am impressed.
I too am an engineer. I have a degree in physics. And I have made a telescope mirror which only requires two surfaces, instead of three because it, at least starts out, as a sphere.
You stated that the three plate method can be used with three straight edges to make them all straight. But this is a common misunderstanding of the three plate method and is not the case. The three plate method says that three surfaces will only be able to be in intimate contact, in all three combinations (AB, BC, and CA), AND in ALL POSSIBLE ROTATIONS when all three are flat.
So the reason why the three plate method fails when making straight edges is due to the relative small width as compared to their length. Your straight edge looks like about 2 meters long but only around 50 mm WIDE. If you try to rotate one straight relative to another by 90 degrees, you will have only a 50 mm x 50 mm area of contact. And that small contact area will never be enough, in a practical sense, to allow you to bring the three straight edges to true planes.
The three plate method is applicable to surfaces that are either round or approximately square. Surfaces like the surface plate that you used as a table and a reference. To produce a 2 meter long surface with the three plate method that surface should be at least 1.5 meters wide. Any further deviation from square would make it difficult. This is perhaps why most surface plates are rectangles similar to this.
If the surfaces are not checked in at least two angles at 90 degrees apart, then there are an infinity of other shapes that can be generated. Perhaps the simplest of these is a surface that is twisted in a manner similar to a propeller. Three such twisted surfaces will nest with each other in all three combinations and even when rotated 180 degrees, which could be done with straight edges. In fact, that twist can be very dramatic to the point that the short edges on opposite ends of the long dimension are as much as 90 degrees to each other - or even 180 degrees. A 1 degree or 1/2 degree twist would be all but impossible to see and to detect if they were rotated to 90 degrees about the axis perpendicular to their desired plane.
Now that is straight ! Very impressive work. Thanks for the tips.
Thank You !! I have a new binge watch channel in METRIC !! Windhoek, Namibia
Glad to see that he is teaching the world.
Great content! Please continue, and thanks!
Loving all content from you lately!
Thank you! Love when old thoughts about how things should be designed are being challenged. Great video and great proof of concept...
Keep them coming :)
I found this channel accidentally - I am totally blown away - by far the most competent and complex work in this genre and I don’t say that just becaspuse I have the same lathe and mill
Very informative information here. Thanks for sharing your knowledge. I look forward to watching more of your content.
Awesome video! This and the one about the cutting geometry for scraping carbon steel will be very useful for my next project. Thanks for sharing!
Beautiful construction plan.
Great project. Wise solution and optimisation.
Thank you for this alternative straight edge fabrication method, it is perfect for my limited machine tooling shop. Big thanks to Stefan Gotteswinter for linking to your channel.
Glad if my hints are of help to you. All the best, Alex
Great content. Thanks for sharing.
Thank you for the excellent presentation! So I am taking the plunge to make one (but quite a bit smaller).
Bernhard
Pretty damn good. That you said, it took you a couple of days, well that sort of flabbergasted me. I'd very much have liked to have seen the process.
This is a very original idea! I was thinking of making camelback straight edges but using a bar of cast iron and a piece of steel as a stiffener screwed onto the back of the bar.
Love your content. Really elevating my thinking, and enthusiasm for my own workshop. Thank you.
Very welcome!
Yes Alex! Keep the videos coming
Thanks for sharing. Nice project and explanation. I found your channel a few weeks ago and love the high quality content. Greetings from the Netherlands.
Many thanks, Sir!
I made my way over here because of Stefan Gotteswinter. Alex, I am glad he made the recommendation! Your identification of, solutions to and explanations of the problems you encounter and the ideas you have to solve these problems is enjoyable to watch. I say this selfishly: keep making videos! I want to watch more!
(BTW, the GTP-13 is super-cool.)
Thanks for your nice comment!
Great stuff! I am somewhat of a scraper myself along with the other normal things machinists do. Anyway, I stumbled upon this video... sort of. I saw it recommended on the side bar. After reading a few of the comments I realized that I too have heard Stefan talk about you, but never investigated any further. Too bad for me!
But now that I have found you and will spend some time watching the other episodes you have posted. ( I have watched a couple others already) So far I have very much enjoyed what I have seen. Cheers
That's one good looking straight Edge
Great content and thanks for your valuable engineering tips :) Greetings, Stefan.
As per Stefan's recommendation, Very nice find!. Very fine workmanship for sure.
Many thanks!
Simply excellent.
Such an underrated channel!
Thank you, great video!
Hello Alex -
I followed stephan g's link, and I'm glad i did. This was a nicely done, interesting video, and i look forward to looking at your others.
Amazing the power Stefan has, one word from him and we all subscribe, and glad of it. 👍
You say it, Sir. I am very overwhelmed.
Very helpful. Thanks!
Having a precision level always helps. You can also use a U-shaped aluminum stock and glue sandpaper with different grits into it and use hand force to make it even smoother. Eventually you can reach straightness with micron precision
Well done.
Hi Alex. very interesting build, and great explanation. Subscribed
Thanks!
At one time at work we were re-building a number of crankshaft grinders. These grinders (Landis IW's and Pin Grinders) use a flat and V way with the female half of the V in the base. Fortunately the Pin Grinders we could grind the ways in house. The IW's we had to send out. In order to scrape the V for the Wheel Head one of the crew scraped a master from a section of Way Cube from a Cincinatti Swing Fixture Broach used to perform the initial machining operations on inline six cylinder engine blocks. These Way Cubes are made of hardened cast iron. That was a real bear for him to scrape requiring him to sharpen and hone the carbide blades about every two scraping passes. When done worked extremely well.
Incredible. Maybe one day people will look at this master way cube and say: "This thing was made like the great pyramids. We don't know how."
@@anengineersfindings
Oh I doubt that. The manufacturing plant is now an empty field and anything that wasn't auctioned off has long sense been melted for scrap.
Awesome.
Pretty good accuracy
nice, this is very cool. 2years and it is still strong and no f casting involved - sweet.
Just got the heads up of your channel from Stefan. Just subbed, good luck with your channel.
Cheers
Many thanks!
Great video. You're obviously a smart guy.
Thanks, but women say the opposite :D
@@anengineersfindings It means they like you ! To them, yes is no and no is maybe.
Good information thank you
Thanks Dale, glad you like the content.
An important point in steel geometric equipment making rather than in iron is the stress relief as steel is usually not stress free and not at all after welding , the DIY solution is to vibrate the beam to liberate strees after welding or to make a wood fire and use a master H beam as rest to avoid deformation of the control beam under treatment and let cool slowly in sand or hashes .
I love it when the tried and tested methods are proven to be not the only way to do something. Everything I have read suggested that steel is unscrapable, not dimensionally stable enough and that weldments can never stay precise over time. I think your straight edge proves all of these wrong.
I am also unable to source cast iron for making straight edges but have easy access to mild steel, which is what led me to this video. Based on this, I am going to attempt to make a dovetail straight edge for scraping machine ways.
Thanks and good luck for your dovetail-straight edge build!
Intresting video! How much does it sag on the middle if its only supported on the ends? Did your consider making it from aluminium? What would be the drawbacks in a simular made out of aluminium?
I just subscribed. Good stuff. If you come to Thailand, I am happy to show you my small shop, have a great meal and some beer.
Thank you for your kind invitation, Sir. I always loved the southeast Asian countries and will surely travel there again some day. Looking forward to a possible visit of your shop. Thanks.
Found from Stefan, subbed. Very interesting construction.
You might need a foam piece over your microphone though- that's some crazy strong echo!
Thanks and sorry if the audio is bad.
Servus Fifi,
sehr interessant, danke für´s zeigen.
Gruß, Sepp
Grüß dich Sepp, sehr gerne, hoffe dir geht es gut. Beste Grüße!
Stefan sent me. New subscriber.
Hi great video and project and congratulations on your result of straightness.
I did some crude calculations regarding stiffness. To my surprise I found out that the greater I made the moment of inertia the more the part sagged (theoreticcaly). The added weight of the part when I made it taller resulted in more sag. Even when I gave the part a 20% weight reduction a 1000mm long straight edge would sag 13,2microns when the straight edge was only supported in each end. I chose this load scenario as the sag would introduce an error when a straight edge was laid flat on a surface plate. Did you do any calculations or FEA prior to fabricating your straight edge or am I just overcomplicating things? 😊
Hmm... without knowing your FE-model, the behavior you describe (more sag with greater height) seems wrong to me. According to analytical mechanics, the max. sag of a beam with rectangular (!) cross section between two supports equals =const.*density*(support distance)^4/(Young's modulus*height^2). So a beam with doubled height should have quarter the sag, even though it is twice the weight. Maybe your model had a problem with the boundary conditions? I assume so, since you see from the video that the sag of this particular straight edge is well below 10 microns.
Wouldn't it have been better to heat and quench the 2 beams and center part beforehand in order to harden them, then temper them to something like 30 Rockwell for the perfect balance of stiffness and ductility, and then after welding it all together normalise the whole thing to relieve the stresses from its production process and the welding? I would think substantially increasing the straight edge's stiffness and reducing sag is an important trait, as is relieving the stress induced by the earlier welding?
This is absolutely fantastic! Did you ever find the laser cutting pattern for the fillet? I don't have much in terms of machines, but I have scraped a few old cast iron hand planes to flat and could use a good straight edge!
Thanks! From experience I'd like to suggest to make the straight edge in a size that fits the job at hand. If it is too large, spotting becomes quite awkward. If it is too small, the same problem arises. The fillet sheet is not so critical, to keep it rigid and lightweight, you can follow the design advice in the video.
Thank you very much. Is it possible to scrape a straight edge of this length with only a surface plate of half of it's length?
I only have a 24" surface plate. Also, is there any way to check flatness of something long like a mill table, without a long straight edge? Thank you very much
Making a precise straight edge double the length of your surface plate will (I guess) be very difficult. If I were in your shoes, I'd rather make the long straight edge by the "three plate method". You can check a long surface also by using a precision level. This can be annoying because of the waiting period for the bubble to calm down, but with large pieces it may be the only option.
Great design, well thought and organised planning and execution ! One question: at min 3:10 you say you had the straight edge stored for about 2 years sitting on that rubber; why that ? why not sitting at Bessel points or hung up vertically ?
Vertical hanging would be the best, I agree. But compared to storing on the Bessel points, storing on rubber results in less stress in the straight edge and therefore less potential for warpage.
If i don't have a surface plate but i have a 600mm surface grinder, can i make a 1200 straight edge using the surface grinder?
I presume the girder surfaces are heat treated steel? How would you scrape that? Very interesting project. Loved you mill rebuild vid.
The girder surfaces are not heat treated, however the girders are out of medium strength carbon steel. I used the techniques shown in my "Hand scraping of carbon steel" video. Thanks.
If those pins are exactly round why can't you role the straight edge over to one side which enables you to measure the complete edge ?
Amazing precision. Did you do any kind of FEA modeing when you designed the straight edge?
Thank you! Sorry if I disappoint you, but I used analytic mechanics to roughly set this straight edge's dimensions, no FEA.
No disappointment. I understand how the bending beam equation can be used to calculate stresses and deflections, but with the special cutouts in the web I thought maybe FEA had been used.
Fantastic information, thank you. As far as I'm concerned the best thing about Covid-19 is high frequency video releases from Dr.-Ing. Fillafer. ;-)
:D Your comment made me laugh out loud, Sir, thanks.
wow this is great!!I do own a lathe that i bought from the scrapyard really cheap and i revived it to working condition but it needs scraping on the cross slide dovetails. Is there a eway to make a dovetail straight edge so as to scrape the ways or moglice them?I do not own a surface plate and the crosslide is 60 cm long. I cant justify the cost for the right tools cause i am an amature...
Thanks and good luck for your upcoming scraping job. I haven't made a dovetail straight edge out of steel yet. Maybe you can borrow one out of cast iron somewhere. If not, you may need to think of a way to fabricate one. Sorry, I can't help you with more advice at this point in time.
@@anengineersfindings you gave me enough already!Didnt think that crs can be scraped flat or stay flat!
Hi; you could have set the straight edge across the surface table to give you more surface for testing
Hi, why would hot-rolled have been a better choice? How long did it take to make the straight edge? I am astounded by what you have accomplished. Thank you for sharing!
Thank you, Sir. Yes, as I say in the video, I'm sure long-term wise it will be better to use hot rolled material. Also as mentioned in the video, it took me roughly 2 days to make this straight edge (lasercutting excluded). Thanks!
Hi, thank you for the video. Two questions though. If we have no tools and we want to go for the three edges method,
1. how do we get it close to the ballpark so that we can begin scraping?
2. When scraping / lapping both sides how do we insure parallelism?
Thank you.
I can't really answer your 2nd question, but in reply to the first one; you apply diamond polishing paste to straight edge A and then put another straight edge on it (B) and while making small figure eights you polish/lap them to match, then switch to the other straight edge (C) and repeat it, making sure to continously cycle between the 3 straight edges to lap them to match each other, using smaller and smaller grit polishing paste as you continue. Eventually you'll probably be only a couple dozen microns from straight away, and the last bit you do by scraping or hand grinding with a small rotary tool and sanding drum.
Greatly thank you, Sir, for such an educating vid! It makes precision machining possible for mortals and self-taughts like i am!
I have one question 🙋♂️ regarding the three plates method:
- Robin Renzetti posted a vid about this method suggesting that it will make the 3 surfaces parallel but not straight if the plates geometry is not round as it cant rotate during the spotting ( which is the case for straightedges)
(Please find the link to the vid in the next comment)
Can you share your insight on this?
I am planning to build some straightedges that way longer than my surface plate 😁
Thank you very much and have a nice day!
Here is the vid i mentioned
czcams.com/video/Va4TGDnQqDs/video.html
From 2:30 - 7:25
Hello! Thanks for your kind comment.
As always, there is nothing much to add to Robin Renzetti's information. His comment may not be of much significance in case of straight edges. That is, only if the straight edge is narrow (more like a line) and not wide like a surface plate. If it is wide in comparison to it's length, then a twisted surface result may be possible by using the pure 3 plate method. This can be avoided, however, e.g. if you check the narrow side with a precision level or if you check the diagonals with a twist gauge (as Robin shows). ATB, Alex
3 straight edge (like 3 plates method) gives you a 3 spiral surafaces. Because you can not turn straight edges on 90 degree angle.
That's a pretty innovative solution! You can't buy straight edges here in NZ, and with monstrous shipping prices, this could be a great alternative. Do you happen to still have the file you used for the laser cut steel? It looks pretty straight forward, but if you don't mind sharing it would be a nice time-saver to simply replicate what you have done.
Also, you should do more videos! I've really enjoyed the one's you have done so far.
Thank you very much for your nice comment, Sir. I'll have to check if I still have the file on my work-PC. However, this might take some time, since I am quarantined at another place right now. Greetings, Alex.
@@anengineersfindings thank you! No rush of course, everybody's in the same boat at the moment, so I doubt I could do anything with them at the moment anyway!
Hi Alex, just touching base to let you know I'm still very interested in using your straight edge design. I have a welder now, so in theory all I need now is to find someone locally who can cut the steel, and of course some way of acquiring some of your awesome skills!
It's been a busy time I know, so as I said before, no rush! 😁
Do you, by any chance have the lasercut files?
@@ChristophLehner I don't yet. Hopefully one day :)
Respect! +/- 3 micrometer maximum over 1 meter.
Did you temperature anneal the instrument after welding and before scraping?
No, that's the funny aspect about it.
@@anengineersfindings really? This corrupts all what we have learned from the books. I am curious how the same measurement looks like in 2 years.
Right, it does. But keep in mind: As I say in the video, at the time of filming, this straight edge was already 2 years old. The trick do mostly the tiny fusion tacks.
Hey great video. When you say carbon steel what exactly do you mean?
Plain weldable mild steels, such as SAE1015.
At 13:20 you explain that the spotting pattern changes due to the overhang. I'm unclear on why.
Wouldn't the edge of support act as a fulcrum, and the weight of unsupported steel would pull down below the surface plate, flexing the nearest supported steel UP off the surface plate? My intuition (which is apparantly wrong) insists that the spotting pattern near the edge would be lighter, and very heavy just at the edge of the granite. What makes the pattern more dense instead?
I guess the "flucruming" and the associated spotting pattern you describe will occur if the straight edge is quite flexible in bending. However, then it would not be very suitable as a spotting tool. The more rigid (stiff in bending) the straight edge is, I assume the less you will observe the spotting pattern you describe. I didn't observe such a spotting pattern in case of the straight edge featured in the video.
@@anengineersfindings - Based on what you explained, I will have to rethink how much that an unsupported tool (even a straightedge) flexes under its own weight. I had incorporated the ideas that "everything is rubber" and "when you measure below 25 microns, everything matters", into how I tried to predict errors in setup.
I believe you saw what you said. That means I have some wrong idea, and it seems likely to be the one you focused on - the straightedge geometry is rigid enough that even at such small measurement scale, the effect is not significant.
What accounts for the effect you DID see, the heavier inking near the edge?
fukin awesome
High carbon or mild steel?
you are cute! :) thanks for the video dear sir, vey interesting. :)
Stefan send me.
if you saw how i did this you’d probably cry.
I'm a straight guy, but comparing with this straight edge I'm a gay ! 😂