Geometric Dimensioning and Tolerancing

Just joined after watching the saga of the fabricated frames and “tables”.
I’m a retired engineer with fab experience prior to ASME Certification (GDTP #0274) and a 30 year career in “Dimensional Management”- An engineering discipline that helps part and tool designers design parts and make tools so the parts just fit.
I try to keep up a column in “The Fabricator” magazine (Mike Matusky) and will try to answer questions on GD&T on The Fireball Forum.

Thank you for doing this.

I wrote this just before I watched the Fireball video on fixture tables…
Response to “When is the Right Time to Build a Welding Fixture?” in “The Welder” Sept/Oct 2023

Is GD&T included in the designs of the component parts?

It has been said that GD&T is a contract with inspection but a really good suggestion for production. The basis for this is when parts are inspected according to their GD&T the (datum) features that establish the coordinate system for part inspection are specified in the design. They are a very specific instruction set for the Metrologist or Inspector.

Sometimes a poor selection for the datums results in the tail wagging the dog; an inexperienced Designer “picks” datum features that are easy for Design resulting in measurements being made relative to meaningless surfaces and edges. What should be happening here is the Designer sharing an understanding with a Fixture Designer to select a set of datum features that indicate how the part will be held in the fixture (or in the next assembly/weldment).

A weld (or assembly) fixture can be thought of as a set of locators (GD&T speak: “Datum Simulators”) for each component to hold them all in the single, correct relationship. Some of the locators are built into the fixture; other locators are features of parts already placed in the fixture. Loading sequence matters, and sometimes it’s critical.

  1. The features critical to part fit up in the fixture will be the features setting up part inspection. Parts that won’t fit are first revealed in inspection, and the reasons they won’t fit can be addresses there, instead of in production. Instead of inspecting only for “part-to-print” you’ll also be inspecting for “part-fit-and-function”.

  2. The most important feature relationships will be the relationships between the datum features. Typically, the flatness of a base feature, the angularity (relative to the plane established by the base) of a second feature, and the angularity of a third feature to both the established planes. For some parts, no feature characteristics other than the qualifications of the datum features will require inspection.

  3. Caution: Non-datum features used to locate parts in a fixture MUST be specifically toleranced and should be inspected. Using a feature to locate a part suddenly makes that feature critical even if the designer only saw it as a “parent” holding other much more important features in their correct relationship.

Flexible parts

Which are better, flexible parts or those that won’t flex at all under clamp loads? Ideally clamp loads will not distort parts being clamped and can sometimes be applied to push flexible parts into place.

“Is there anything wrong with parts that just fit?”

Weldment (assembly) GD&T

While we’re talking GD&T, Does the weldment drawing place appropriate emphasis on the right assembly features? Consider a frame made of square tube with gussets in the corners with holes in them to attach a screen panel inside the frame. How should these parts be located?

Set aside for now the decision (wisdom?) of whether the tube corners are mitered or butt joints. Let’s just consider the four mounting holes for the screen.

We could clamp the square tubes against fixture pins surrounding the frame, then clamp the gussets into the corners and weld them there. See Figure 1:

Note that if the setup in Figure 1 is chosen, short side tubes that are too long WILL NOT FIT.

Consider it might be nice to find this out BEFORE welding.

Alternatively, we could pin the holes down in the fixture and clamp the square tubes to the sides of the gussets. See Figure 2:

Again, short side tubes that are too long will force gaps into the weldment.

Or, we could locate two perpendicular tubes and clamp everything else against them. See Figure 3:

At this point consider the difference it makes how the corners are designed. Butt joints, as illustrated here, create a different set of challenges than mitered ends. IN EITHER CASE, consideration should be given as to whether tight joints, with the parts setting the final dimension would be preferable to small gaps (that can be welded over with impunity) that allow the fixture to set the final dimensions.

And the CAD jockey who “designed” the weldment has no clue. In that world everything is perfect.

The first method tends to make the frame the right size while the second tends to put the holes in the right pattern. And the third may be the easiest to load. Which is most important? GD&T, used correctly*, will tell you.

Pull Quote: *Using GD&T correctly does not simply mean without violations of applicable rules. GD&T can be on a drawing, and not incorrect, without taking advantage of its ability to express things like “These four holes are used to attach a screen with a matching pattern of holes” or “The size of the outside of the frame is critical for fit and function”. Or it might even say both, which would indicate a need to be very careful with both component dimensions AND fixturing.

Consideration should also be given to which is more accurate, your fixture elements or your component dimensions.

3-D Printing for Fixture Elements

Many welding tables include a precision grid of holes to locate fixture parts. Exactly what (pin, clamp, angle, etc.) goes where is important to document. To take this a step further, 3-D printing can be used to create dedicated fixture parts that hold specific components in specific locations; two stub pins for the weld table on the bottom, a pin or clamp pad in a specific location on the top. These dedicated fixture parts can be stored in minimal space for reuse, or re-printed as needed.

Geo-Set + Re-Weld vs Single Weld Operation

One way to avoid weld distortion is to only tack parts together while they are located in the fixture. After the weldment geometry is set by tacking, the weldment can be removed and the welds completed. This geo-set and re-weld process often allows easier access to back side welds and may be necessary to avoid damage to 3-D printed fixture parts.

Specified weld sequences may also be critical to prevent welds pulling the parts out of location. Document weld sequences as well as where pins, clamps, and pads are located on weld tables.

There is much to consider in the design and construction of weld or assembly fixtures. GD&T can be used to highlight critical relationships between parts, fixtures, and assemblies.


Mike, I believe your article takes the discussions around Jason’s fabricator/fixtures videos to a new level. For me as an amateur, it certainly sheds a bunch of new light on a topic that I was only vaguely aware of. Thank you for contributing the article, I hope it will inspire those of us who are unfamiliar to broader horizons and improved processes! On one level it could be simply appreciating the skills/experience that should back the various roles you call out: for those of us winging it with the rudiments of 3D CAD and a welder, to think for a bit about how to properly design parts and processes is golden. On another, it may inspire one to delve deeper into the engineering and learn something of the language of GD&T.


@mfmatusky Excellent work on this thread. I also am hitting my 29 year mark working in the metrology field and you did an excellent job explaining the challenges of designing, dimensioning, fabricating, and inspecting the parts.
I joined this forum to give some input from my experience in metrology and manufacturing regarding the welded table video but I look forward to seeing more of your insight here and hopefully learning from you to fill in some gaps in my knowledge.

Soryu- So glad to be providing useful knowledge! As I moved toward retirement I realized that the skills I had acquired were in many cases hidden or dispersed in many Product Lifecycle Management (PLM) plans. In writing post retirement I am hoping to bring them to the forefront where more folks developing and managing products can see and appreciate what often unsung contributors have long been doing.

Something else that comes into play, with setting GD&T limits for production, is Interchangabilty and Rreplacement. Much of what I work with has a huge IR component to it’s fabrication. We have to be able to take any part fabricated under an IR requirement and show it will fit in the same location on any installation with absolutely zero modification to the part or location. We also have parts that are what are known as IR-A. The basic part is IR, but can be hand altered to fit the exact installation, where the part is not one that is designed to be normally removable. An example would be a riveted down skin panel that a forms one or more edge to a removable panel or door. Makes the design guys and gals work a bit harder, but makes it much simpler for the production people.

Yes. GD&T can be used to indicate how the part installs; what locates it in place. If a panel is located by an attachment hole and slot it doesn’t make a lot of sense to measure it with respect to two edges.
Using GD&T doesn’t by itself assure that mounting / locating features are used as datum features though.
Also to the point is that GD&T does a better job of setting BOUNDARIES in 3-D space that represent the mating installation geometry. As in: “If the part extends beyond here, fit-up can not be assured”.

I went looking for a GD&T reference. The ASME Standard is both expensive (Like ~ $130 depending on revision) and text books can be similarly expensive. I came across this, 166 pages with both explanatory text AND diagrams (some taken directly from Y14.5).
Best of all, it’s free.