Notched rectangular wood beam 2

[TahoeEng]

Hi, I'm hoping somebody can shed some light on a "field condition" of a 5.25"x11.875" PSL header. It has a holdown from above so it has an all-thread through it with a bearing plate on the bottom of the beam. The contractor notched in the 5"x5"x1/4" steel bearing plate, the notch is about 2" deep x beam width x 7" length. The beam span is 6.5' and the notch is 2.5' from the bearing end. The beam has a full length 1200 plf uniform load and 5800 lb point load directly over the notch (2.5' from end).

The header is in a 9' plate height wall. Siding is on, window is installed (only retrofit available seems to be on the interior face).

I know per NDS this condition is not allowed per Figure 3C - Notch Limitations for Sawn Lumber Beams. And I am extremely concerened about the stress concentrations and potential failure due to the notch.

I called TJI and they said they treat their PSL beams the same as sawn lumber for notches.

1) does anybody know of a way to determine the capacity of the notched beam?
2) does anybody know of a way to alleviate the stress concentration (possibly using a steel side plate sistered to the beam?)

Thanks in advance for any help!

 

[dik]

The Canadian Code has a method for calculating notch strength on both the tension and compression fibre side, but it is somewhat complicated. I tried cutting and pasting it from the *.pdf file I have for it, but the formulae don't paste well; the formatting is lost. I can fax the relevant parts if you like. Post your fax number to dikcoates@alpha.to

 

[RG88]

TrusJoist MacMillan does not recommend notching the beam in the fashion you describe (according to their literature on PSL beams).

Assuming that I understand your post correctly, the notch is on the bottom (tension side of the beam). For a simple approach, will the beam work as a 9.875" deep member (11.875"-2" notch)? I haven't run the numbers from your post. Basically assume that the bottom 2" does not contribute to the strength of the beam, but double check the shear due to bending parallel to the grain at the notch depth for the full depth beam and check the shear (tension) perpendicular to the grain at this same depth from the point load (you don't want to exceed the allowables to avoid splitting potential). Because the beam is full depth on both sides of the notch, you cannot discount the extra depth in its entirity, especially in regards to shear due to bending and shear perpendicular to the grain. This is a simplified approach that must exercise some judgement in the factor of safety.

Unfortunately the Timber Construction Manual does not cover notches in the span, only at the end bearings. There is a way to calculate this like dik mentioned, but I cannot locate the reference for notch sensitivity.

In your case tension perpendicular to the grain can create the potential for splitting.

A moment resisting splice can be made to the sides of the beam with steel plate. The Timber Construction manual covers some of the methods to calculate this type of repair.

 

[TahoeEng]

Well, RG88 and dik, I'm still looking into the strength of a beam with a notch.

dik: I'm looking forward to seeing what the Canadian Code has to say about this issue.

RG88: I don't think the beam will work as a 9.875" deep member since as you stated the tension perpendicular to grain increases the potential for splitting. I looked through the Timber Construction Manual and couldn't find a repair method, could you point me in the right direction?

I did speak to Truss Joist and the local rep/engineer said that he didn't think that the location of notch was critical and they used a factor of (d/de)^2. Where d=depth of beam and de=depth of beam less the depth of notch. I'm not sure how I feel about this. I'm thinking it may be okay since shear parallel to grain and moment (which I think induces the tension perp to grain) to not have maximum values at the same location. Shear max at ends and moment max at midspan, typical of uniform loaded beams. Yet, I think the value should be less than that calculation produces.

I have been doing some "tests" with strips of paper. I've been cutting "paper beams" approx. 2"x8" and placing on a smooth flat surface and ripping them. I've been putting square notches and tapered notches in the "tension side". It is amazing how much easier the paper rips with any type of notch.

Well, client and contractor have been put on hold until I can get a better feel, or can justify a value of the beam.

I appreciate your help and look forward to any further input.

 

[RG88]

TahoeEng

I am not clear on the (d/de)^2, it is my understanding that this type of correlation is used for end bearing notches. I was wrong about a specific (to your case) detail in the Timber Construction Manual on tension splices. In the Appendix to AITC they have some dos and don'ts for connections some of which discuss splitting issues arising from fasteners and splices on glulam arches. I have the fourth edition.

One method to consider is to develop a tension side splice with a steel strap of sufficient length and strength to allow for fasteners with a moderate to high factor of safety. What you are trying to do here is to hold the notch in place and carry the tensile stress to the other parts of the tension side of the beam. Watch dowel bearing strength, pre drill the holes to reduce splitting, stagger the bolts a bit, and remember that each bolt will not, in practice, carry the same load (thus the higher FS).

You may have to break the news to your client that the window may have to be removed to place a bottom tension side splice if the side mount variant will not work.

 

[boo1]

Check this link:

https://www.aitc-glulam.org/shopcart/Pdf/aitc_tn_19_ may_2002.pdf

down load for free, I hope you find it helpfull!

 

[TahoeEng]

boo1: Thanks for the link. Some interesting reading...yet it only addressed notches at the bearing end. I just called AITC and hopefully their tech support team can help additionally.

RG88: Yep, a tension splice sounds like a viable option, although at that point I might just make them replace the header and move it up. They have room since it's a 9' plate. Thanks for the input.

dik: Much thanks! I appreciate all your effort in getting me the info. I look forward to seeing it.

I'll keep you guys posted, at this point it's still up in the air....

 

[rlflower]

One more thought:

Putting aside all the technical jarogn for a moment, let me address a simple concern. When you measure, do not simply measure the depth of the notch but sure you include the full depth of the cut.

The reason I say this is the main reason why notching is widely dismissed as unacceptable; contractors tend to overcut the notch to make it easier to cut out the un-wanted material. Cored holes are therefore more preferred.

 

[TahoeEng]

Well, folks, I think I have exhausted my search for a "canned" solution for the notching.

dik: Thank you so much for forwarding the sections of the Canadian Code. I assume you were referring to Section 5.5.5.3, but once again, only notches at supports are addressed.

rlflower: Point taken on the overcuts associated with notching.

I did a site observation and luckily the header has a LVL rim joist above and the wall is furred which will allow a "un-notched" member to be sistered to the notched member.

I also did get a return call from AITC, they do not have a method to address notches within the middle 1/3 of the beam.

Unfortunately I can not report a resolution on the matter for anybody's future reference, except, do not allow any notches other than what your code allows near the supports!

Thanks again for all the input. And of course, if anybody does have a solution.... please post it.