Ok, I thought the links described a fatigue test quite well.
1) you need a test article (got that), that needs to be conformed to some set of engineering. You may want to install s/gauges to collect data at either control points (to verify the loading and your FEM) or at points where you have done analysis.
2) You need a load apparatus ... a hydraulic jack is the start of this, you also need a "whiffle tree" to distribute the load over the test specimen. Many other details like how to support the test article (will it be grounded, fixed to the ground at a point, or will it be "flying", with complete balanced set of loads and reactions) how to "tare" the test specimen (as weight is not part of the loading ... you could count it as part of the reaction but not typical practice). Many questions !
3) you need a test spectrum (I've said this repeatedly). Yes a test spectrum is different load levels, yes 1g and 1.8g many be representative; but how many load cycles to apply for 1 hr of flight, 1 flight cycle, whatever you choose.
4) Again repeatedly I've asked for certification or for company experience ? If for certification you need to talk to your local certification agency, and they can help but you'll need more technical support than you have demonstrated to date.
5) I think you've said this is for a UAV. I suggest that current spectrum generation tools are not appropriate for this type of airplane. I suggest that you measure the spectrum of loads your plane reacts to but doing flight test (with a loads monitoring system installed), collect data for several thousand hours to be somewhat confident. I'd also use this loads monitoring system on delivered airplanes, to gather more data, to verify the data you've collected.
6) I think you need professional help !
"Wir hoffen, dass dieses Mal alles gut gehen wird!"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.
Many articles do not give the fatigue test process, so I asked for your help.