Shaft alignment spec on small machine 1

[ShaggyPE]

I am involved in the design of a device that utilizes a rotating shaft driven by a T-head gearbox to two driven devices on each end of the shaft. All three of these devices are mounted to a common plate. Note that the shaft rotates slowly. Approx 100 rpm. The shaft diameter is approx 1/2"... plate ground flat... This is a small benchtop machine and the devices weigh less then 20 lbs.

The discussion point is regarding the "need" for specifying a method to align the two driven devices to the T-head gearbox (and shaft). One recommendation is to "simply" torque all mounting fasteners as a group once fully assembled. Effectively using the shaft to align the driven devices.

Ultimately the debate is:
Does there need to be any specifics on the face of the drawing to ensure the shaft alignment process takes place as described above? Or are there industry standard / millwright standards etc that can be assumed will be followed.

I realize I am posting this in a drafting forum amongst engineers. I expect an overwhelming "we need to define everything," and "never assume." However I am playing a little bit of devil's advocate and curious if the task doing this sort of assembly should carry the expectation that qualified assemblers would just know to ensure alignment via whatever background training (qualification) they have. Similar to tightening lug nuts or head bolts.

I have spent the better part of the day trying to find examples (online) of machinery assembly drawings to make a case either way. However as would be expected, companies don't have their proprietary documentation readily available.

Looking for feedback from other industry professionals. Thanks.

-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[drawoh]

ShaggyPE,

This is more of a mechanical engineering problem than a dimensioning and tolerancing one. How much mis-alignment can you tolerate? How capable are your assembly people?

You can use your shafts for alignment, but I would expect things to shift when you tighten stuff down.

--
JHG

 

[ShaggyPE]

Thanks for the reply drawoh. I agree that it really isn't a dim and tolerance issue... I am looking for drawing notation feedback though... The question regarding capability or the expectation of capability and how much hand holding should be on the drawing is really the question. The parts are fairly high precision relative to the interface with the ground plate. Any elevation variation due to the heights are within part tolerance, however any lateral variation can be viewed as due to assembly misalignment.

Ultimately, I am looking for either examples of drawings that specify a process for something like "tightening head bolts" or examples where that type of process definition is omitted because the installers are expected to already know the process.

If this assembly process were "more complicated" or more frequently used by varying devices, we would likely create a standalone internal document to define the process. The drawing would typically contain a note stating "Align IAW 1020863." But where do you draw the line on defining a process for a task that is viewed by many as an expected understanding.

I realize the conservative approach is to include a note of some sort. But is it bad that a note was not present? Are there examples of a similar process where a note was not present? Is there overwhelming evidence for the process being defined?

-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[drawoh]

ShaggyPE,

If you can assemble the thing and get it working, there is a way to document the process. If the
instruction is simple, it can be a note on the drawing. At some point, you need the power of a word processor to write stuff down. Specifying a minimum level of skill by the worker is done all the time in manufacturing. Skilled workers are of course, more expensive.

I am just picking numbers out of the air here. Take the case that you cannot align your shafts to any better than .001". When your machine runs, your shafts will be strained [±].001" as they rotate. What is the resulting stress, and what is your fatigue life going to be?

Can you make your shafts longer?

--
JHG

 

[ShaggyPE]

So further information. The device has been built. There was no notation on the drawing for the alignment process. When it was assembled, turns out there was a bind in the shaft. Time was spent testing/troubleshooting before the bind was identified (stepper motor stall). Ultimately loosening of the supports and re-tightening them in a balanced fashion cured the problem. A note has since been added to the drawing to define the process. Finger pointing has ensued. The question that I am being tasked to investigate (yes, tasked) is if it was acceptable to create the drawing in the first place with the process note omitted. Ideally, if I can find some examples either way, it would allow me to close this task.

I have found some larger machinery recommendations (facility pump install and alignment type stuff). However those are typically installed by millwrights who do have a high level of training is the area of alignments. At that level (I could be wrong), I believe the process are not defined. Perhaps just the goal. I say perhaps, because the bit of research I have done leaves that answer vague.

I am not finding examples of what other companies or industries do for their assemblies where a process slightly more complicated than a torquing lug nuts (single part with several fixed fasteners).

Input certainly appreciated.

-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[drawoh]

ShaggyPE,

Documenting an/the assembly procedure is, to me, a basic DFMA practice. If you have a plan to get the thing assembled and working, then we can expect the thing to be assembled and got working. If manufacturing can find a better way, that is nice. If manufacturing can find no way...

If I were checking your drawings, I would want to know how your shafts got aligned. Either you show me flexible couplings, accurate enough and manufacturable tolerances, or an assembly procedure. Another possibility is that your company routinely does this, and your assembly people are accustomed to dealing with it.

Do these shafts ever break?

--
JHG

 

[ShaggyPE]

Does your opinion change if the designer offered his assistance on the build of this first article and the MFG Engineer declined said assistance?

In hindsite, having a note (process) is the right answer... but how wrong was it that the process wasn't defined.

I am an intermediary in this activity. Not the designer, not the assembler.

-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[drawoh]

ShaggyPE,

Is this production, or a one-time-only assembly. In production, IMHO, full documentation is part of due diligence. If this is limited production or one-off tooling, assembling the first one and showing everyone how to do it should be sufficient. We are moving away from engineering here, and into the realm of office politics. How well do the designer and the manufacturing engineer get along?

--
JHG

 

[ShaggyPE]

It is the first article of the production device. We are certainly in the realm of office politics... and I'm stuck in the middle. The designer and the manufacturing engineer do not get along. I like the statement: "In production... full documentation is part of due diligence." So how big of a miss would you consider the lack of an alignment process? The drawing was drawn by a drafter (not the designer), it was checked by a third person and reviewed by the designer and another engineer... and reviewed by a manufacturing engineer.

Being in the middle, is the reason I was hoping for concrete examples one way or the other. Still hoping if anyone can provide. Not looking for actual drawings, just further opinions or examples of how their industry works.



-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[drawoh]

ShaggyPE,

When you quoted me, you left out "IMHO". It's important.

I was recently laid off from a place that generally does not give a s**t about stuff like that. I am now in a place that is fanatical about GD&T and DFMA and design checking. I am enjoying it thoroughly.

Quite a few years ago, I designed an adjustable mirror that consisted of three screws and stiff springs. I positioned the screws 90[°][ ]apart so that the adjustment would be fairly orthogonal. I moved the screws as far apart as I could, and I placed a note on the drawing describing the range and precision of the adjustment. The range is controlled by the solid height of the springs, and the compression that generates minimum acceptable clamping force. The precision is a function of screw pitch and separation. When the platform was redesigned for a new laser, someone moved the screws closer together and out of the orthogonal configuration. This made the mirror way more difficult to adjust. The resulting platform was left in production in that state for over ten years.

If they are stupid and they don't care, keep your resume up to date and network like crazy.

--
JHG

 

[handleman]

This really depends on your company's expectations of the people doing the assembly work. I worked in custom automation equipment for 15 years. In our environment, assembly technician knowledge was expected to be at a high level. When assembling any rotating, sliding, etc. motion, they were expected to have the responsibility for careful assembly, alignment, and motion confirmation. Unless there was a specific need for alignment to a non-obvious datum or measurement, there was never a specific assembly instruction on something as obvious as "rotating shafts should rotate after assembly".

 

[3DDave]

I would consider initial lack of alignment instructions as not worth mentioning. It's obvious no one involved spotted the need during the initial design, the early design reviews, or the prototype stages. Whoever is in charge of managing the project didn't budget for a useful stress analysis or tolerance analysis. I would hazard a guess that any request for help along those lines would have been ignored.

Seems like resume time for all involved.

 

[SDETERS]

Since the Shaft if rotating slowly, could you put in a flexible alignment coupler to take the cost out of precision alignment of your gear box to the shaft? The designers and engineers designing the box, should have done some tolerance stacks to determine if the shaft will bind or rotate freely.