Dumbest thing an engineer ever did 11

[RontheRedneck]

A while back in the thread I started about engineers being wrong someone asked me to post something like this.

We sent out some trusses that needed to be field spliced. Huge 3/4" plywood gussets.

The framer called me and asked "What's the tolerance for the cuts on the plywood gussets." No one had ever asked me that before. So I told him I'd call the engineer who drew the splice detail and ask.

I made the call and asked the question. The conversation went something like this:

There are no tolerances.

What do you mean?

There are no tolerances. The cuts have to be right.

The cuts will be made by humans with hand-held saws. There have to be tolerances.

No, there are no tolerances. Tell them the cuts have to be right.

When they make parts for the space shuttle there are tolerances. Everything has tolerances.

No, there are no tolerances.


I eventually figured that the conversation was about as painful and unproductive as beating my head against a tree. So I gave up.


In my entire career that is the most ridiculous conversation I have ever had with an engineer.

 

[phamENG]

That is rough. A simple -0,+1/2" would have gotten him what he wanted...

 

[dik]

Great response... and would have ended the discussion quickly.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[RWW0002]

I would have asked him what the tolerance is on his tape measure.. Ha!

While definately frustrating for all parties, at least this "dumbest thing" doesn't result in loss of life or any kind of real failure..

 

[3DDave]

First establish a datum reference frame and use the profile of surface geometric characteristic control.

My tolerance for wood work for a hand sawn project is within 1/32 of an inch.

 

[Smoulder]

"No one ever asked before"
So do there need to be *specified* tolerances? From engs perspective, useless question dealt with with minimum fuss. Think about that.

 

[phamENG]

I think it also exposes a big issue in our industry, though, and one we should probably all try to rectify. For instance: there is no codified tolerance limit for wood construction. There is for steel. And most engineers working with steel are familiar with it. But when it comes to wood...an engineer can go their entire career working in the design of wood structures and, if they're lucky, never have to worry about tolerances. Part of that is that in most cases the framer can make it up as he goes and fix things on the fly.

I didn't learn about it until I left consulting and went to work on site at a steel mill and had to play mechanical engineer for a few designs. I learned about tolerances from the machine shop real quick. But I imagine those guys had a similar feeling to Ron when I walked away, though I'd like to think I was more receptive than the engineer in Ron's story.

 

[Factoter]

Heard this anecdote from an engineer somewhat the reverse scenario:
He was between jobs and so picked up a gig with a remodel contractor through an acquaintance; the contractor agreed to hire him right off, offering a sub position installing a kitchen countertop:
Twelve feet of new countertop (this was back in the day, so probably Formica).
This engineer agreed to the job, feeling confident it was in his wheelhouse, so to speak;
Before digging in, he asked what the tolerance should be for leveling, to which the contractor replied emphatically: 1/64’’ it must be within a sixty-fourth front to back, side to side.
After very careful analysis, deep metrological cogitating and nitpicking, our poor engineer friend quit the job after a better part of a week in anticipation of a paycheck.
Thems the ropes.

 

[IceNine]

"Structural engineering is the art of molding materials we do not wholly understand into shapes we cannot precisely analyze,
so as to withstand forces we cannot really assess, in such a way that the community at large has no reason to suspect the extent of our ignorance"

-Unknown

Of course there's a tolerance, and it probably isn't very small.

 

[kipfoot]

this seems less about being wrong and more about not understanding what you were asking for some reason, like being a non-native english speaker.

 

[dik]

When I used to build furniture, dovetail clearance was within a few thousandths...

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[canwesteng]

Well a few thou too big, but it can always be 1/32 small. But with a dovetail if it comes in just a bit too big, you sand it down to fit. Should have sent the engineer the gussets and some sandpaper and the gussets and asked him to give them back when he was happy.

In any case just glad this wasn't a post about one of my projects haha

 

[JoshPlumSE]

When they make parts for the space shuttle there are tolerances. Everything has tolerances.

This reminds me of a story an engineer (let's call him Evan) told me early in my career. He was working on a project related to military cargo jets. He had to design a platform that would support some workers and equipment as they constructed some portion of the jet. This was something like 25 feet in the air and the footprint of the equipment couldn't be more than 15ft square.

The client gave him a specification that the equipment couldn't move more than some absurdly small tolerance like 5 micro inches. He told them the specification was wrong. They didn't listen and told him that's what it was. Eventually there was a meeting between him, the client, our project management, their project management and such.

Evan: "I've run the numbers and the only way I could make that tolerance is if the support stand were a single 15 ft square block of concrete. Even then we just BARELY make the tolerance."

Client: "Maybe we should have an ARCHITECT take a look at this for us."

Evan: "Correct me if I'm wrong, but I don't think you care what color the support structure is do you? That's the only input an architect could have about a support structure like this. I think you mean you want a structural engineer."

Client: "Yes, of course. Forgive my mistake. Let's have a structural engineer take a look at it."

Evan: "Excellent. It just so happens that I am a structural engineer. Can we just go back to whoever wrote this specification and ask them for more details about why the deflection tolerance is so strict. It just can't be correct for this type of structure."

Eventually, the client did this and realized that the equipment was mounted to the jet's air frame directly. So, the deflection tolerance was completely irrelevant to the structure that would support the workers.

 

[rb1957]

"A simple -0,+1/2" would have gotten him what he wanted ..." ... WAY too optimistic. The "engineer" has repeatedly said "no tolerance".

What you should've done was tell the vendor this, -0,+1/2", and tell the guys installing them "they may come a little large, trim to fit".

It would've been funny if the dimensions were slightly off (and needed to be slightly oversize).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[RontheRedneck]

phamENG said: "...there is no codified tolerance limit for wood construction."

There are some tolerances related to trusses. But I can't think of where to reference them off the top of my head. If I can find the information I'll post it.


kipfoot said "this seems less about being wrong and more about not understanding what you were asking for some reason, like being a non-native english speaker."

Nope. I've known this particular engineer for ~30 years. It's about being bull headed, thinking they know everything, and having not ever gotten their hands dirty in their life.

 

[Eng16080]

An engineer I know was involved in a project at an airport where a lot of catch basin grates were needed. The price they got from their normal supplier was too high, and so they determined they could save some money by having the grates custom fabricated. The recess in the frame for the grate was 2 ft square. The specifications sent to the fabricator by one of their engineers failed to mention any tolerance, and simply required the grates to also be 2 ft square. So, both the recess in the frame and the grates were fabricated to be the same dimension. If I recall correctly, in the end, none of the grates fit!

 

[phamENG]

Ron - I think TPI-1 has tolerances for truss manufacturing. But, from experience, how many of us engineers actually read that before specifying a truss?

I'm referring to wood construction more broadly. There are books on the subject that try to bring together general industry practice, but their rules of the thumb and non-enforceable unless written into a contract.

 

[KootK]

Meh. This is an entertaining anecdote but, functionally, there has to be empathy and understanding on both sides of most technical discussions.

this seems less about being wrong and more about not understanding what you were asking for some reason, like being a non-native english speaker.

That's my guess as well.

I'm sure that the engineer sized the gussets to respect edge distances on the fasteners and all that jazz. So my answer to that question probably would have been:

Give me more gusset than I asked for, everywhere that I asked for it. Now leave me alone, I've got real work to do.

Which is, of course, the non-robot way of saying -0,+1/2".

 

[KootK]

With respect to wood tolerances, I encountered some difficulties myself in that regard as expressed in this thread: Burying Steel Beams in Wood Stud Volumes. Annoying.

For the benefit of any interested parties, I also have to excellent books on tolerances. The first is a super handy reference that I quickly use to zone in on the tolerances applicable to unfamiliar spaces. The second is just downright entertaining. The tolerances that apply to the jet engines on airplanes is gonzo. And the stakes are high.

 

[Aesur]

This reminds me of a story an engineer (let's call him Evan) told me early in my career. He was working on a project related to military cargo jets. He had to design a platform that would support some workers and equipment as they constructed some portion of the jet. This was something like 25 feet in the air and the footprint of the equipment couldn't be more than 15ft square.

Oh man, this brings back some memories of a similar situation, client wanted a pool on the 5th level of a wood framed apartment over a 3 story podium. Pool engineer said maximum deflection of all members was 0" so they can use a concrete shell instead of the more expensive steel shell. Numerous conversations occurred, including how I would not support a pool on wood trusses and beams and insisted on using steel and that they need to give a realistic deflection limit that they would not budge on, so I did the obvious, I told them it would be a 8 story solid concrete block from foundation to pool on the roof. They quickly realized maybe it's worth the extra cost of a stainless steel shell, but still refused to budge on being supported by wood. Many more conversations and my asking them to sign some paperwork saying they directed me to support on wood and accept all liability and case dropped, pool deleted from project. They didn't have the budget for the pool anyways, the architectural design caused the budget to be 2x their funding in the end.