Check for wire wound resistors, they are most likely to be capable of a high
pulse of current. But check for the manufacturers specification on what
they have been tested to tolerate.
There used to be several sources of these, but I haven't used any for years.
Of course a insulated bobbin with copper or iron wire could be used, if size
isn't critical. Wind it as a single or double layer to get the heat out.

On 24 Nov 2003 22:50:40 -0800, (Jason Hsu)
wrote:
It's part of a design for a T/R sense circuit for a noise cancellation
device.
I won't bore you with too many details. One thing I noticed in a
design I'm looking at is that 1W resistors could be subject to as much
as 50W of power apiece during the time it takes for a relay to
respond. This response time is 7msec. The overall duty cycle will be
low (well under 1%).

Can the 1W-51 ohm resistors handle this 50 RF volts 0-peak (about 50W
PEP) for .007 sec? 50W over .007 seconds is .35 Joules. .35W for 1
second is also .35 Joules, which a 1W resistor should have no trouble
handling. Can the resistors be damaged during that .007 seconds?