Help resolving guardrail manufacturers calculations. 1

[shacked]

Architect wants to use a proprietary guardrail system for residential. I downloaded their calculations but something just doesn't make sense to me when I review their calculations. I have attached them with red clouded areas that are pertinent.

Analyzed is Risa with a 200# pt load at node N20 in the Z direction they get a 295ft-lb moment at the base node N16. If this is correct then that is a big difference between the typical single post static analysis, with no load sharing.

Also, on page 9, their base plate calculation is kind of confusing. They are resolzing the 295 ft-lb moment at base plate/post into a point load(I guess a 3" moment arm) then multiplying that force by 1/2" to get a moment of 708 in-lb??? This doesn't make sense. If the moment arm is 3" and the width of the square base plate is 3.5", then what is the 1/2" moment arm?

I would think that the base plate should resist the full 295 ft-lb being developed between all 4 bolts.

Thanks for your input

 

[SteelPE]

First, I am not a RISA guy at all.

At first glance it does seem confusing to me too. I have run plenty of similar calculations and I don't typically use load sharing. They also appear to be ignoring the 50 plf loading requirement in any direction (maybe because this is a one and two family dwelling???).

I don't like the fact that there are no details to go along with their calculations. It makes it incredibly difficult to follow along with the analysis.

With regards to the moment on the base plate.... there is a moment in the rail that is suddenly driven into the base plate. The base plate will have a moment diagram that is both positive and negative at the post (I really don't want to draw a picture sorry). I am guessing the bolts are located 1/2" from the center of the post and they are calculating the moment based upon this distance, but without a detail it's hard to tell.

 

[dauwerda]

I personally have no issue with load sharing - the model is showing that it happens, so take advantage of it.

Handrail and guardrail systems shall be designed to resist a single concentrated load of 200 lb (0.89 kN) applied in any direction at any point on the handrail or top rail to produce the maximum load effect on the element being considered and to transfer this load through the supports to the structure.

This doesn't seem to make any indication to me that the guardrail can't be designed as a framed system.

The base plate calc is first finding the tension in the anchors assuming a rigid plate. This is found to be 1416 lbs. Then, the bending in the plate is determined by multiplying the distance from the anchor line to the face of the column, which is 1/2". See my sketch below:

Personally, I believe the tension in the anchors should have been determined:
Without the rigid plate assumption (conservatively assuming the compression acts at the compression face of the column, or less conservatively as somewhere between the face of the column and the edge of the plate. - this would result in a tension of 1770 lbs as opposed to 1416 lbs and a bending in the plate of 885 in-lbs
or
With the rigid plate assumption with the required thickness of the plate being determined with the bending moment using the 1" moment arm between the face of the column and the edge of the plate. - this would result in a bending in the plate of 1416 in-lbs which is greater than 1310, so a thicker plate would be needed - edit, it results it 1180 in-lbs (their calc was incorrect) so a thicker plate is not needed (with these design assumptions).


**Upon further review and me running their numbers, they actually used a moment arm of 2.5" in the initial calc to find the tension in the anchors, 295*12/2.5 = 1416.** So they used a compression force halfway between the face of the column and the edge of the plate.

 

[shacked]

Ahhh, thanks dauwerda that makes sense now.
I was assuming that the plate would bend between the bolts. This was obviously a wrong assumption since the post is welded at the center and it can't bend in the very center.
AlthoughI still don't see how they determined the t=c=1416.

Moment arm between bolt holes = 2inches. Therefore the T=C=295*12/2 = 1770Lb. They used 3 for some reason and still wrote the incorrect resultant down since 295*12/3 = 1180Lb.

 

[jjl317]

Perhaps a silly question, but have you contacted the engineer who stamped the calculations?

 

[BAretired]

I checked the base plate using Yield Line Theory assuming three yield lines would form in the locations shown. Looks like the base plate is adequate.

BA

 

[shacked]

Thanks All.

JJL317, I knew someone would ask that. No I have not. I assumed it would be easier to post here 1st before I tried to contact them through the company.

Makes sense now. Now I have to determine is I can make the anchorage into concrete work since the contractor wants to epoxy in the anchor bolts. Considering that the bolts are pretty closely spaced, at first glance it doesn't look like the anchorage will work, but this isn't something I will need help with.

Regards all.

 

[CANPRO]

I'm all for load sharing between posts and I have looked at this closely in the past. As SteelPE noted, there is usually a requirement to design for uniform loading on the guard rail which would negate any load sharing effect. And if you're going to go with the load sharing approach to get the best economy out of your posts/baseplates then you'll likely need a different post/baseplate detail for the last post, which won't benefit as much (if at all) from load sharing with adjacent posts.

Are the calculations meant to cover all baseplates or just a typical baseplate for an intermediate post? As mentioned in the paragraph above, the end post will not benefit as much from adjacent posts.

 

[Ron]

Load sharing depends a lot on the top rail. If you have a continuous top rail, the moment in the base plate will be governed by lateral deflection in the top rail and the end connections. If you have a joint in the top rail at a post, that post gets little help from load sharing except for the integrity of the connection...sometimes good, sometimes not so much. I do engineering for a local stair and rail manufacturer and have run into this numerous times.

For rails with no end connection at the top rail, it is difficult to get baseplates and posts to pass due to high moments.

 

[KootK]

It is all but impossible to make typically proportioned rail systems work without the load sharing. You might consider downloading a free copy of the manual shown below which describes some of the common assumptions used, provides some examples, and offers some tabulated information.

 

[SteelPE]

So it appears as if the 50plf requirement isn't required for one and two family dwellings, so you can share the load between adjacent posts.

Interesting that they used a wood load factor of 1.6 for the 200 lb load. I understand what they are doing, however, the code does not state that this load is a short duration load. The requirements for rail loading fall under the live load section of the IBC and of ASCE-7/. That being said, I may try to use this load factor in the future (I have always used 1.0).

 

[r13]

For a 2 span railing, what is the design load for the middle post, based on load sharing and no load sharing?

 

[BAretired]

You could either make an educated guess or perform a frame analysis. The end posts would take 200# each, so I would design the middle one for the same.

BA

 

[r13]

So, for 2 - 6' spans, the design load for the middle post is 200#, or 300#, or else? I suggest to read this article. Link

 

[strucbells]

Interesting that they used a wood load factor of 1.6 for the 200 lb load. I understand what they are doing, however, the code does not state that this load is a short duration load. The requirements for rail loading fall under the live load section of the IBC and of ASCE-7/. That being said, I may try to use this load factor in the future (I have always used 1.0).

Isn't this for a steel or aluminum railing, not a wood railing? My understanding is that the load duration factors for wood are related to the specific material properties of wood that allow for increased resilience to short term loading. Therefore, these would not be appropriate or typically used for another material.

EDIT: Good point dauwerda, I missed that last page of calcs and misunderstood SteelPE's comment here.

 

[BAretired]

r13,

No thanks, I'm not too interested and I'm not sure it applies in my neck of the woods. With 6' spans, I would probably choose 300#, but you could put up an argument for even more than that because of continuity. Probably, none of the posts will ever feel the load specified in the code.

BA

 

[dauwerda]

The calc package included information for fastening the railing to wood with lag bolts. The 1.6 factor was used for determining allowable withdrawal values of lags in wood.

 

[r13]

Do not take the railing design lightly, as it is the source of tragic events on the news - people falling from stairs at partying, especially on residential buildings, or club houses.