Yee, I remade the prints for you to help avoid confusion. You can just straight use them idc, but I’d recommend using them for reference as you remake your own. I quickly made them so it’s possible mistakes were made.
1 Like
You never stressed to the fabricators that you needed the parts to be as accurate as possible. Yes, the plans have specs attached, but they are simple parts without an obvious need for accuracy. It’s generally assumed if something highly accurate is desired for an unconventional/simple build, something verbally will be said to the fabricator to indicate that.
1 Like
i agree . most fabricators don’t expect to have to make stuff like this to such tight tolerance. and i bet most customers accept what they weld up . i have made large steel fabricated molds like 18 ft x 6 ft x 2 feet deep and it is a real challenge and you have to have a strategy for dealing with distortion and minimizing it. those molds make 100 fiberglass parts that all have to align to each other sitting on beams. so a deck.
your 3 fab shops should have looked at the drawings and questioned the tolerances on the drawing telling you they can’t meet them. likely they don’t due tight tolerance work . it’s also the nature of fabrications that everyone in the industry accepts distortion as inevitable. you should do a video of you fabricating the parts and how you would do it. i suspect you would do the T setups first then weld the pieces in the 4 corners. even then you would deal with the tubing itself not being a perfect square tube,
As a fabricator I have learnt never to just says it’s easy because even simple looking items can sometimes be a challenge. I treat every discussion with a customer as a learning session not only for myself but also the customer.
So the first question I would discus is the all-round fillet welds and mention that the side welds are not fillets and are actually a Bevel Groove that is naturally formed by the radius of the tube.
I would also ask if the welds are to be seen or will they be hidden, because this will determine the weld process I will use, because if my welds are going to be seen I want them to not only be functional but to also be pleasing to the eye.
As far as using a fixture table, they are a great idea, I wish I could afford one, however my bench is a nice flat piece of 40mm thick plate, that I got for a slab of beer. Looking at the tolerance and knowing how welds distort things I would probably tack weld the two square frames together back to back then put the legs on to reduce what is going to happen to the square frame and also reduce any twisting.
Great video look forward to the next one
This post is perfect. Especially the comments regarding, “Is this hard to make vs. I’m concerned with meeting tolerances”. I felt the same way. I think the test, and the shops chosen, were destined to fail, and if Jason had specifically stated that there are tight tolerances (even if he was bluffing and said it was part of an assembly, so the tolerances are critical), the shops would have either declined the job, or paid more attention to quality.
2 Likes
Most fabrication shops would scoff at a customer who had expectations that a fabrication might be measured on a CMM. The costs associated with that would explode! If the dimensions and tolerations are functional it might be best to thicken the tube walls and mill to maintain dimensional and form tolerances.
There is a misconception that GD&T equals “precision”.
I’ve had this argument many times with machine shops that lob that criticism that GD&T is some how “space age” or “means it’s atomically precise”. It’s not.
GD&T was created to allow the MOST variation possible and still have a functioning part, and describe those dimensions unambiguously.
The reason GD&T was made was so torpedoes of multiple manufacturers would all fit in the same launcher hole in the 1930’s. Because, it’s kind of inconvenient when you’re trying to fight off submarines and you find out your torpedoes don’t fit your launcher.
It’s one of the big reasons it’s so popular in automotive, and military. Because it allows you to make the part as cheap as possible and still function.
Some notes on your drawing, which I would give to a junior engineer presenting this to me. And remember, this is critiquing the drawing not you 
.
We’ll use ASME Y14.5 Y14.1 Y14.3, and ISO 2768 for reference.
- You don’t need a drawing for each tube in the assembly, that’s the job of the weldment cut list.
- Title Block - quantity isn’t needed here, normally would be handled in the BOM in an upper level assembly or at time of purchase. Need a revision letter, can be omitted if using a revision block on the top level page. Scale should only appear in the main title block area to avoid confusion (Y14.1). Missing material call out and finish call outs. In your “unless otherwise specified” section the tolerances should become smaller as the decimal place increases, shown below. Missing angular tolerance. When making an assembly with multiple sheets, make the sheets as a pack so the fabricator knows there is more than a single sheet. Each sheet having a “sheet 1 of 1” implies that each part is independent of the main assembly. Also generally a good idea to add a “interpret per ASME Y14.5” or whatever standard you’re using so other shops can apply the same standard when interpreting. I’d probably also add a block or note that says “dimensions apply after welding too” for good measure.
E.g.,
.XX ± .05
.XXX ± .025
.XXXX ± .001
Use ASME Y14.5 §1.6.2b for reference when dimensioning in inches as well. “A dimension is expressed to the same number of decimal places as its tolerance. Zeros are added to the right of the decimal point where necessary”
Aside: Sometimes I like to use the ISO 2768 tolerance ranges since it’s based on basic size of a part.
- Good use of the third angle projection and projected multiview. technically we could remove the second side view as the part is fully described with only the top and one side view.
- Dimensions: reduce decimal places down to the nearest applicable tolerance required (see Y14.5 excerpt from above). Enclose the 9" and 9.125" in brackets to indicate they are for reference(Y14.5 §3.3.8). Otherwise the dimensions over constrain the tube. Extend the centerlines for the 10 and 10.125 dimensions to show the dimension is for both sides. Move the 20 from the side view to the top view as they are redundant in the side views.
- Weld annotations: “TYP” is a deprecated notation, use quantity instead. Bring the welds for the square from the other sheet to the top view of this sheet. This may be more preference, but I would keep a V-groove instead of a square weld.
- Some basic GD&T to fully describe the part. I am making some assumptions here, like the square base is the A datum, and the corner is the B and C datums, and since I don’t know the function I’m giving it a LOT of tolerance. Excuse my crude sketches for GD&T. First I would add a flatness tolerance of 1/16 and assign that to datum ‘A’. This means that the flatness of that face must be within a 1/16 boundary and acts as the first datum (using three points of contact). I would next add a parallel requirement of the leg faces. This block is interpreted as all the faces of the legs must lay within a 1/8 boundary from datum ‘A’. The “CF” means continuous feature so it applies to all legs. In the top view I would change the 10 and 10.125 to basic dimensions, and use the position tolerance block. The block is read as the tubes have to be: perpendicular to A, up from B, and over from C datums, within a 1/16" boundary.
I’m sure I’ve made mistakes here, but in essence adding those GD&T callouts to the drawing unambiguously describes the tolerance the part must lay within. It also helps the fabricator understand how the part should be manufactured. They need to have a table that’s at least 1/16 in flatness, and they need to be able to measure the height of the legs to within 1/8. And that they can add hard stops to only two edges of their fixture table.
I also replied elsewhere, but GD&T is NOT to make a part super precise, it is used to allow the MOST tolerance possible and still have a functioning part. As well as unambiguously describe the tolerance the part must lay within. Anyone that tells you GD&T is “space age” or “aerospace” or “super precise”, has no idea what they’re talking about.
Also sorry for the rant 
1 Like
Personally, I’m fascinated by the number of people here that argue he should have specifically called out and emphasized the tolerances when dropping off the plans. They were right there, written down, and not exactly hard to spot. I don’t see why he should have to specifically point out the tolerances any more than any other measurement.
If the fabricators didn’t spot them right away, they could have called him before getting started building the product after examining the plans in more detail, and turned him down if they couldn’t meet his specs, or wanted other discussions about the tolerances, like negotiating them down, or simply stating they had no great way to measure some of these (like flatness.)
3 Likes
It would definitley be a good exercise to create a revised print. When in doubt leave notes, even if it means creating a second page!
I agree. I have honestly come to appreciate both sides of this argument, having dealt with this in the past. But I learned early, as a fabricator, ask lots of questions and tell the customer what to expect… and let them decide if they want to use you. For example: “if you absolutely require flatness, it will require considerable adjustments in how we make your part, and the cost will be reflected in the quote.” I can’t count how many times someone took the time to make very nice drawings, but had no real concept of what needs to happen to achieve a noted requirement. So shame on these guys for not asking the important questions, AND shame on Jason for expecting Joe, the dude down the dirt road who welds all our tractors back together, to make the parts right.
Let’s face it, Jason’s products are, IMO, the finest welding fab tools on the market. He has been exceedingly innovative and has worked his ass off to make top quality goods. He’s in a tough spot, trying to compete with the big dogs who have been around for decades. Like Wilton vises. They are NOT special… they were just the first to produce something better than anyone else. Jason has to fight an uphill battle to show that, dollar for dollar, you are better off buying the hardtail. Kurt vise was the go to for decades, but their time has come and gone, and their new products are kinda crappy. Give me a Parlec any day of the week, but I bet most folks won’t know the name, so they keep buying Kurt. Jason needs to find better ways to show how good is shit is without demeaning the other products. I mean, this isn’t politics.
2 Likes
Are you going to build one to your specifications did I miss that already?
2 Likes
That is the incorrect tip on the CMM. You should be using yhe tapered one. That one will not properly calibrate the way its displayed in the video.
I would very much like to see you yourself weld these frames within spec. Also it seems like you are choosing shops which you know will not meet your standards and then continue to put a microscope to their work to make yourself look better. Maybe try taking it to a machine shop? and not a welding shop if you’re looking for precision where you know you wont find it.
2 Likes
I think it would be interesting to see your approach to solving this with two or more methods/toolkits. In one example, show and test how you’d do it with no limits in your shop. In another do with much more basic tools, similar to what your local shops had. Maybe a third where (assuming you used some heavy machinery in step 1) do the “bare bones” but with a fixture table.
How close do you get in each? Could you design a process to meet your own tolerances? What are the pros/cons of each?
1 Like
I would also listen for what questions they ask me. I’m a software developer, not a welder, but I always make sure I understand what’s important to my customer. Who’s going to be using this for what? Will they be able to fix or work around problems themselves?
If this is a subcontract because the client is in a rush, then maybe they’ll have to do fitting in the final assembly anyway, so which dimensions are gonna make that step easier or harder for them? What are they expecting to have to fix? Should I leave some cuts long for final fitting? Now I know which measurements to prioritize.
If this is the final product, what is it for? Are the tolerances real, or just there because you had to put something?
As the customer I would be concerned if I left the shop and they hadn’t made a single mark on the drawing before I left.
(Sorry about deleting and reposting. First time here and just figuring out how these threads work.)
1 Like
Your test for twisting was fun, but a bit flawed. It works for the first set of frames, but for the tables it doesn’t actually isolate twisting from bowing. Because of the bowing you already know is there, a part with zero twist would give you exactly the same result.
BTW, I love watching your videos. They’re like precision porn.
1 Like
I think the way you demonstrated tolerances in the video was incorrect. When you press one corner to the table and measure opposite corner, you’re potentially doubling the actual error in the object. If the spec asks for 1/16" accuracy for each corner, then every corner is allowed to be off by 1/16" in any direction when the object is in the static orientation that minimizes the errors. That is, if the center is touching a flat table, both ends can be off by 1/16" at the same time and still be within the spec!
The same goes for the legs: the spec doesn’t give any accuracy to twisting a leg on frame, only asks for 90 degree angle without any tolerance definition. You cannot pretend that it means that it must be a perfect 90 degree angle because no such exists with a physical object.
You cannot really complain for tolerances that were not specified in the drawing, can you?
That said, I agree that all the errors you measured were obviously errors in manufacturing but complaining about non-speficied accuracy demands is not cool. It would be insane to manufacture to one thou unless specifically asked for it.
Also, when you write
Fractional dimension tolerance: +/- 1/16"
in the spec, does that mean that the dimensions for the edge that has been labeled as 20" has no tolerance defined at all, because it was not defined with a fractional dimension? Even if that was not your intent, could the manufacturer have reasonably misunderstood your spec?
I have always loved problem solving and learning how things work. The way you problem solve is such a great mindset to be in. I have 12 years of experience in this line of work and over that time have seen, learned, and taught a great number of skills. After watching your second video on how other fab shops deal with your tasks I felt it’s time I give my 2 pence worth. My response will not be exact to your question but more to how I feel about this situation so the problems I see are improper practice and lack of final inspection. I thought the 3rd workshop would have had something close after hearing about using clamps and seeing their table having a PFC frame so the datum could be better than the rest but sadly still failed. I truly feel the workshops you are using have people who have gotten into the mindset of “that’s good enough”, “we are not building helicopters” or “this is how it has been done for years”. you should try presenting this task to a bigger company to see how they differ. I would love to have the equipment you do my life would be so much easier but at the same time would become more boring without the fight to get things right. I have thought of my process for this task but have spoken for long enough if anyone would like to hear what these processes are I would be more than happy to explain but that is it for now. finally, note your videos have helped me explain things to other fabricators who are stuck in their ways I have learned from your videos also and am most grateful for your time and efforts thank you.