Barrier is only designed for TL loading but then again you shouldn't have to design a temporary barrier, assuming you're using a DOT approved barrier. You may have to design the anchorage of the barrier to the deck if your state DOT doesn't have detail or if you're not permitted to use another DOT's detail.

I am using an approved DOT barrier and it is bolted to existing concrete deck slab. The question is do I need to use the AASHTO LRFD Bridge Design Specifications for 1/2 of existing bridge that is only there for 6 months?

I wouldn't as it doesn't make any sense. It was built in 1951, why check it for a 2017 LRFD Bridge Spec? If they want an HL-93 load then I would check it with ASD or maybe LFD with the newer truck.

I had a phased deck replacement a couple years ago on a bridge built in the early 70s. I think I did check the effect of running a single lane of traffic on 2 girders, but I checked it according to the Standard Spec and HS-20 loading as that is what it was designed for and how it was Load Rated.

I wouldn't use LRFD.

Thanks bridgebuster and OSUCivilEng. I had meeting with DOT Bridge Dept and said I didn't think the 4 bolts barrier would transfer moment to the deck and they said that it would most certainly transfer moment to the deck. They said this must be analysed as extreme event using LRFD and also add multiple presence factor to live load of 1.2 since it is a single lane albeit a temporary one for 6 months?

It's not an unreasonable request. At the end of the day, the client/owner/DOT is going to get what they want -- or you'll find another way to protect the live traffic that will be running on your bridge.

A traffic impact load is not quite like a seismic or wind lateral load -- even if you argue that there is less probability of it occurring in 6 months (which may not be valid in a construction zone scenario), the optics of a structural failure are very, very bad. That translates to high risk for the DOT.

Whether you check LRFD, LFD or ASD for the existing structure shouldn't make much difference for a concrete T-beam structure. I'd expect you might actually get some benefit from LRFD (although it will be a longer calculation).

Like OSU mentions, if the current bridge was rated at less than HL-93 before replacement began, you could probably use that to argue for a reduction of the HL-93 requirement, which would decrease the deck slab demand some.

I agree with the DOT -- most barrier details would have to transmit moment to the deck slab to be effective.

----
The name is a long story -- just call me Lo.

You definitely need to design the barrier/anchorage for whatever TL level the Client requires. As mentioned above, typically a standard load tested barrier is used so design of the barrier itself can be skipped. However, one should ensure that the anchorages to the existing deck are satisfactory for a reasonable test level. Typical detail here is a threaded rod through the deck bolting down the barrier. As for the structure itself, I recently completed a staged bridge rehab on a frankenstein of a structure - we load rated the bridge and used every Clause and allowance we could to reduce the calculated live loading. If I remember correctly, there is a formula to calculate the live load factor in the load rating section which is based on a number of site/structure specific parameters... Our calculated factor came down a bit from the generic factor provided in the design section... could be of help.