Hi Rahimi, At it's core, LRFD is just a way of splitting up uncertainty. Rather than designing with one single safety factor, we attach different factors of safety to our estimates of capacity and demands depending on the nature of the failure mode and nature of loading. As for decreasing the resistance, failures that we can accurately predict and have "desirable" behavior (like ductile failure modes) do not need as much safety margin as failures that are more uncertain or have "undesirable" behavior (like brittle failure modes). As for increasing the loads, some loads are relatively predictable (dead loads) while others are less so (live loads). LRFD allows us to tailor the safety factors to achieve an overall and consistent probability of failure. Regarding elastic versus inelastic design, this gets a little weird. LRFD design is normally intended for elastic structural analysis to find demands, but regularly incorporates plasticity into the capacity (one example - per AISC the flexural capacity of a steel beam uses the plastic section, not the elastic section). The various factors of safety and the prevailing methods of design mean the structure typically remains elastic under "service loads", that is the expected day-to-day loads - the structure will exhibit some plastic behavior under "factored loads", but should not reach it's full plastic capacity at any location. There are other methods for analysis, such as plastic analysis, where some sections are assumed to be fully utilized and plastic so long as they remain ductile. This can increase the overall capacity of the structure - ductile, indeterminate structures are capable of reaching full plastic moment capacity at some locations and then redistributing the load elsewhere as load continues to increase until a mechanism forms. LRFD isn't normally intended for such things - for such types of analysis, you generally would want to know the sectional capacity accurately (without factors of safety) to know exactly how load redistribution works. Other parts of structural design, such as overstrength factors for seismic design, complicate this picture even more, but that's far more than can be addressed in a KZfaq comment. Thanks for watching, and thank you for your question!