Portal frame deflection? 2

[swazimatt]

Hi am am having difficulty finding information on what is an acceptable deflection amount for the apex of a steel portal frame for a warehouse i am designing.

I have done the calcs and have members that work, but am not sure if the deflection is too much(110mm for a span of 17.3m)?

thx

 

[asixth]

AS1170.0 which is the structural loading code for Australia and New Zealand gives a suggested serviceability limit state criteria of span/300 for permanent loads. For a 17,300 span the deflection limit would be 57mm. So going by this criteria you would need to double the stiffness of the rafter. Depending on magnitude of the design wind, you should also check the deflections of the rafter under serviceability wind conditions, we assume a span/150 or 30mm limit for rafters that are supporting a ceiling and internal partitions (the 30mm can be relaxed for a warehouse because the roof structure will be exposed). The check of serviceability wind should be done for wind loads only because the deflections will be relative to the pre-existing dead load deflections.

The general procedure I take and encourage others to take when designing steel roofs is as follows:

1. Position shear splices at points of inflexion (a portal frame would require a moment splice if you are splicing the rafting, maybe you can fabricate and transport a 17.3m rafter to site).
2. Design for serviceability conditions.
3. Design for ultimate conditions and put fly-bracing where it is required for uplift conditions.

If mechanical equipment is being supported on the roof, try an get it positioned as close to a support as possible. Being a warehouse, I think you will only be getting some air conditioning units on the roof which wouldn't weight more than 100kg (220lbs).

 

[BAretired]

The Canadian standard recommends maximum story drift of h/500. For a height of 17300, that would be 35mm. I think 110mm is excessive.

BA

 

[apsix]

For lateral deflection of industrial buildings (no gantry cranes)due to wind I typically use;
h/150 for steel clad
h/250 for masonry clad

I also take advantage of nominal column base stiffness when calculating serviceability.

 

[asixth]

Maybe I have misunderstood this thread but I thought when mentioning deflection at the apex of the portal frame that swazimatt was referring to vertical deflections?

However lateral deflections is still a very important check and must not be ignored.

apsix

how do you include the nominal base stiffness. Are you considering the baseplate and foundation system and given a rotational stiffness at the restraint. This will also have a small influence on the apex deflection because it is stiffening the column and redistributing the forces for mid-span back to the knee joint.

 

[swazimatt]

Thanks for all the replies. My question was actually serving two purposes, first being to find the allowable vertical deflection for the apex of a portal frame.

But what instigated it is that i was asked to look at a design&build proposal for a carpark roof. Attached is a photo of a similar roof that is spanning 16.9m and is made up of IPE200 sections, column bases are pin jointed.

the new roof is 17.3m (400mm longer) and the D&B contractor has insisted that it should be done using 203x133x30 UB sections. There is alot of politics and unhappiness behind this design and the project manager feels that the 400mm should not make a difference and feels that the D&B contractor is just trying to use up some UB's that he has in stock (the IPE is 22kg/m and the UB is 30kg/m)

I analysed it using prokon (i made up the curve using 2.5m sections with fixed joints) and found that the Mr of the IPE200 was exceeded (for the 17.3m portal) at the column/rafter junction. So i have analysed it using the UB for the column and first 2.5m of rafter and the lighter IPE200 for the middle rafters

I have not analysed the existing structure but am sure that it would have more than the 100mm deflection that i get using the UB/IPE connection.

 

[ooox]

L/300 for permanent loads seems reasonable.

Don't forget if deflection is governing your design you can camber the rafters. Similarly, if you're finding you have large eaves deflections, pre-setting the frame can compensate for these.

 

[asixth]

I wouldn't let the contractor try to influence your design by saying "we have done that elsewhere" and likewise I wouldn't be designing something on the basis that another consultant has done a similar design.

Looking at the photo that you attached it looks like you are going to have some things that work in your favour.

1. Minimum loading. The loads that you will have will consists of the purlins and metal sheeting and whatever electrical is up there for lighting. I wouldn't be suprised if the deadload (excluding the selfweight of the rafter) exceed 10kg/m^2 (2lbs/ft^2) depending on the thickness of the sheeting.

2. The curved rafter will negate the illusion that the structure is deflecting, possibly allowing a 100mm deflection for deadload. If the design is governed by strength, increasing the length of the span by 2% will increase the flexural stresses by 4%, so this may be a driving factor that makes you increase the sizes.

3. The open roof structure is not going to have the same magnitude pressures co-efficients that you would get over an enclosed building, reducing the amount that wind will govern the design.

 

[apsix]

asixth

I agreed with your vertical deflection limits and therefore didn't comment on them.

Nominal base stiffness I use is from BS5950;
5.1.3.3 Nominally pinned base
If a column is nominally pin-connected to a foundation that is designed assuming that the base moment is
zero, the base should be assumed to be pinned when using elastic global analysis to calculate the other
moments and forces in the frame under ultimate limit state loading.
The stiffness of the base may be assumed to be equal to the following proportion of the column stiffness:
a) 10 % when checking frame stability or determining in-plane effective lengths (I don't use this);
b) 20 % when calculating deflections under serviceability loads.

There is guidance in the attached AD097.
Note that BS5950 increased the nominal stiffness from 10% to 20% is it is mandatory to use 4 bolt baseplates in the UK (for OHS reasons). I'd use 10% if you have a 2 bolt baseplate.

 

[swazimatt]

thanks for all the info guys. I think you will be hearing more questions from me on this forum from now on

 

[asixth]

apsix

Thanks for the reference. I will read into it when I get the oppurtunity. I am always looking for ways to make my designs more efficient.

You mentioned that for OHS reasons that all base plates are required to be a 4-bolt connection. I have been assigned the task of developing a safety in design database at my work and I am always looking for things that maybe a safety issue on-site. Can you please explain why the 4-bolts are required for safety in the UK. Does this requirement also apply for column base plates for low loading situations, I have designed many base plates with only 2-bolts.

 

[apsix]

asixth

The requirement for four HD bolts is for safety during erection, so the columns can be free-standing, up to a point. I believe the rule doesn't apply if the column are below a certain weight.
There is no such requirement in Oz (yet) that I'm aware of, and I also use 2 bolts for smaller columns.
Also, refer to attachment.

 

[OzEng80]

Excellent references apsix! That extra 20% could make all the difference.