There is no special test that I know of, at least in D1.1. Multiple pass fillet welds are just that, 2 or more fillet welds overlapping each other. If your welders can pass the 5/16" fillet, which can be more difficult than it seems, they should be able to do 1/2" fillets. It sounds like they may be trying to carry too much metal for the pass. If they adjust the power (amps and/or volts) and the travel speed, while staying within the parameters allowed, they should be able to do it. More practice should solve your problem. If your WPS won't allow reductions in parameters, you may need to write (and qualify if not prequalled) a new WPS.
CHG

Figure 4.36 if for the qualification of welders for fillets only. Once the welder has passed all of the acceptance criteria for the fig. 4.36 test plate then table 4.9 establishes the limitation of variables for the welder. The welder is then qualified for unlimited plate thickness and conversely unlimited fillet weld size.
You seems to imply in your question that the welders technique is poor for making multipass fillet welds. What do you mean when you state that the "welders are not able to do it". What does your welding procedure tell them to do? Please provide more information.

Dear Miguel P,
According AWS D1.1:2000 Table 4.9 you can use for almost unlimited plate thickness - structural steel. In case of 5/16" (8 mm) if you want to make fillet weld in one pass you should use 0.045" wire with following parameters:
I=220-250A, U=22.5V, v=220-270ipm, gas mix = 18%CO2+Ar.
For 1/2" fillet weld please adjust parameters (e.g.: I=300-315A, U=24.5V, etc...). The welders can do it, train them.
To test the specimen you must break the test piece, and make Macroetch (AWS D1.1:2000 Table 4.9)

Sincerely yours Geza Burik