Transverse member in Grillage analysis 1

[TewitC]

Link
I started a topic to ask about transverse member property in grillage analysis and how accurate the method is. Please click the link above to see the discussion.
Some questions I still have are:
1. What is the difference between the result from plane and downstand grillage? My understanding is that downstand grilling taking into account the composite action, but can I use the result from downstand directly?
2. Is there any criterion on when and where the downstand should be used?
3. Which one is more common in general?

 

[r13]

See P3 of the liked paper. Link

 

[IDS]

I think some of the comments in the other thread are quite misleading. A plane grillage analysis does take account of composite action, because the longitudinal members are given the properties of the composite beams. It is true that a downstand grillage will give more accurate results because it includes the effect of variation in the NA depth, and it models transverse distribution more exactly, but the difference is not large. I have attached a conference paper I wrote in 2004, comparing different types of grillage and FEA models.

In answer to your questions:

1. As noted above, the results of plane and downstand grillage analyses are usually quite close, provided they are compared properly. The plane grillage directly gives the moment on the composite beam, but for a downstand grillage you have to add in the moments due to the axial force in the beam and top slab.

2. I don't know of any firm criteria. Anyway, rather than using a downstand grillage I would suggest using plate elements to model the top slab, and beam elements for the main longitudinal beams. This has advantages in extracting local wheel load moments in the top slab, and is useful for skew decks, where grillage analysis can get a bit messy.

3. I'm not sure now. When I wrote the paper (2004) plane grillage analysis was certainly standard practice. It's worth comparing the two on structures you work on anyway.

Incidentally, I did my first work on bridge deck analysis in the UK, starting in 1975, and even then use of grillage analysis was widespread, and it has been standard practice for beam and slab type bridges everywhere I worked since then (UK, Australia, Turkey, SE Asia). It always surprises me that the USA seems to be so resistant to its use.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[Settingsun]

IDS, did you restrain the slab nodes against rotation around the vertical axis as Hambly says to do? I would have thought that requires the slab to bend in plane rather than only shear, but that is what he says it prevents. Have I misunderstood? Figure 7.7 in the 1976 edition, Fig 7.15 in 1990 edition.

 

[IDS]

Steve - I don't recall, and the paper isn't clear on the issue. It says the grillage analyses followed Hambley recommendations, which suggests I did apply the restraints as recommended, but I can't find the actual files at the moment to check.

I did a search on Hambley Bridge Deck Behaviour Figure 7.7, but the only relevant thing that came up was a link to some guy who had written a paper for the Austroads Bridge Conference in 2004 .

I think the point is that because the continuous slab is modelled as separate beams, bending about the vertical axis will be very much greater in the model (if not restrained) than it is in the actual structure, but it seems to me using a very high bending stiffness (as he also recommends) should fix that, without restraining the nodes against rotation.

I'll try and find time to have a closer look at how much difference it makes (but very busy at the moment).

It's another good reason to use plates for the top slab rather than beams anyway.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[r13]

The linked paper compares 3 methods on bridge deck design, not an excellent writing, but may provide you with some insight. Link

 

[r13]

I read a tutorial material which says "A grillage does not analyse in-plane loading so any longitudinal or transverse restraint will not be modelled." Does this answers your question?

 

[IDS]

I read a tutorial material which says "A grillage does not analyse in-plane loading so any longitudinal or transverse restraint will not be modelled." Does this answers your question?

A plane grillage doesn't generate any in-plane loads, but a downstand grillage does, so the question is how those are handled, in particular rotation about the vertical axis.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[r13]

Yes, IDS is correct. But there is very little design/modelling information on the downstand grillage method in the open space, at least in the US. This paper compares results from downstand grillage method to several other analysis methods. Link

 

[IDS]

Yes, IDS is correct. But there is very little design/modelling information on the downstand grillage method in the open space, at least in the US. This paper compares results from downstand grillage method to several other analysis methods. Link

I guess you didn't look at the paper I attached to my first post?

That's my paper from 2004.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[r13]

Well, I made a fool out of myself. The tech world is rather small.

 

[TewitC]

Sorry for the late reply, been very busy lately. Thanks for your comments.
Just one question still isn't clear: In downstand grillage, can the results be used directly? Or do I need to manually combine the action between transverse and longitudinal element?

 

[IDS]

In a standard downstand grillage the longitudinal members have the properties of the composite section, and there are no longitudinal members in the plane of the top slab, so in that case the results for the longitudinal beams are used directly.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/