Tapping hole through two welded plates 1

[dmertz]

I have two pieces of carbon steel A36 plate that were welded together to make a thicker plate. Is it acceptable to tap a hole through both of them at the same time to get the bolt thread engagement I need? Or should I replace with a single thicker piece of plate? Any concerns about corrosion at the gap between the plates after we tap through it?

 

[chaoscreature]

When you have two plates welded together there will always be corrosion between the two plates, you probably just wouldn't ever see it. It's acceptable to tap through multiple layers, but to say whether or not this is a good idea for your particular situation is impossible without knowing more details. If using s single plate is an option would that not make things easier (no welding)?

 

[dmertz]

Thanks for the response. My particular application is the following:

I am mounting a centrifugal compressor to a baseplate which we are designing. On the baseplate design, I have to weld a mounting pad (1" thick plate) to the top of an I-beam. This plate will be machined flat.

To minimize the thickness of the pad, we would prefer to drill and tap (for the compressor hold down bolts) through the pad and the I-beam underneath. This would give enough thickness for engagement of the hold-down bolts.

Otherwise, I can beef up the pad by going to a 2" or 3" plate. However, that would mean reducing the I-beam height and selecting a different beam.

 

[kingnero]

cannot you use a bolt and nut arrangement?
I wouldn't worry too much about corrosion, as you'll use plenty of cutting oil when threading a 1" thick plate...

 

[chaoscreature]

What is the size of the bolt that you are using?
I am assuming the use of nuts on the back side of the plate is not allowable because the nuts would be right where the beam is?
Is the process: weld the plate to the beam, grind the plate flat, then tap and drill through both the plate and beam? This sounds perfectly acceptable at first glance assuming your hardware selection is appropriate and the thread engagement length is adequate. Remember the A36 will have a lower strength than most hardware steels. This is a good resource for threading problems:

http://www.fastenal.com/content/feds/pdf/Article - Screw Threads Design.pdf

 

[dvd]

I would never tap a hole through two (or more) thicknesses. Use one thickness of plate sufficient to accept the fastener and drill a clearance hole through the beam flange, or maybe drill a clearance hole through the plate and beam, then weld a nut or tapped plate on the underside of the beam flange. This would give you a longer fastener which would be better for holding tension.

 

[dik]

One of our engineers that is a contracts administrator was trained as a machinist... he said it is not a good idea... and can often result in broken taps and damaged threads.

Dik

 

[desertfox]

Hi
I have mounted pumps and motors on frames with welded pads to a beam and tapped holes straight through as you describe and never had a problem.
i wouldn't weld a nut to the underside of a beam although I have seen that done also, the reason I wouldn't do it is the heat from the welding can affect the nut material properties.

 

[chaoscreature]

dmertz,

Could you please let us know what the expected loading is on the bolted connection and what size the hardware is. That would help us help you with this dilemma. I am picturing a 3/4" bolt or something around that size, in which case tapping through 1" plate + the flange would be fine, but if this is a 2 1/2-4 thread or something huge than that would change things a bit.

 

[dmertz]

The bolt is M20 and the load is a static 15 kN with a dynamic +/- 3 kN. Sorry, my tool is in SI units. So your 3/4" estimate through 1" plate is a good estimate. We usually use a 2" minimum plate and drill and tap through that.

I was wondering more about the method of fastening, though. I had opened another thread to discuss this topic: Link. So if you have any comments on whether a threaded fastener connection is better than a through bolt connection, please answer them there.

 

[chaoscreature]

An M20 bolt, grade 8.8 has a tensile proof load of 110 kN.
Using the formula:
Engagement Length = 2*Tensile Area / (0.5*pi*(Major Diameter of Screw - 0.64952/#threads))

Tensile Area = 0.40816 in^2
Major Diameter = 0.816 in
# Threads = 10.16

I get 0.691 inches of thread engagement. Now this is assuming that the thread and the steel plate have the same strength.
If the yield stress of the bolt is 92,000 psi and the plate is standard A36 steel with 36,000psi yield than the minimum strength to ensure that the bolt will fail before it rips the threads out is: 92/36*0.691 = 1.765 inches
I think these numbers are very conservative. Kamax fasteners recommend 1.0*d for grade 8.8 in A36 blind holes, which means your threads are fine.
So, because your load is only about 4,000 lbs AND because most of the load is carried by the first few threads I don't see a problem with this arrangement even if you didn't engage any threads into the I-beam. Of course some other items come into play here such as pre-load and fatigue and whether or not people are standing under this compressor (ie is this bolt a single point of failure?)

Nut-Bolt vs threaded plate each have their pros and cons. Nuts are easier since you don't need to tap parts.
Best of luck on your project!

 

[moltenmetal]

We do this sort of thing repeatedly on our modular plant structures and have NEVER seen a failure at such a connnection due to corrosion, freeze-thaw delamination of the plates or any other cause. It is mostly done to increase effective thread engagement in bolted connections to HSS (box tubing) by locally thickening the wall- where the option of going with a thicker wall to start with is a non-starter.

We do not find this practice to be a cause for tap breakage. The causes of tap breakage we see are the use of taps past the point where they're dull (this is the most common), drilling the wrong sized root hole, or an over-aggressive or careless operator.

 

[rmw]

Now that we have the facts, look back at your OP. The question as asked, if it pertained to say, two 6mm plates welded together with a long span for example has a different meaning when asked about doubling 1" plates. Two thin plates with a wide span between the welds would be difficult to tap through as the tap as it tried to enter the second plate would cause the first plate to deflect as it 'climbed' the tap. Not even an issue, however with 1" plates.

rmw

 

[dinjin]

Is keenserts possible?