The same may apply at the global ship level if the ship has been gener-ally over-designed against the global loads (e.g. because the local load requirements were more severe). These aspects can be dealt with by using the panel and global utilisation factors Up and Us defined as follows. panel undamaged for the value allowable max.load design extreme under panel damaged on load max.= p U structure undamaged for the value allowable max.load design extreme under structure on load global max.= s U Normally the allowable strength reductions Rpa and Rsa for panel and global (ship) strength will be set equal to the respective utilisation factors. There are two main additional considerations that may make a further strength reduction acceptable. The first of these is to reduce the loads (relative to the original design) by restricting the operational envelope. This leads to a lowering of the utilisation level in that the extreme design loads are now decreased. There are two distinct ways to restrict the operational envelope: • Restriction in relation to season and geographical location, as defined in the International Load Line Convention and used for defining service restric-tions in classification rules. • Application of a more restricted speed-heading-wave height relationship than that within which the vessel was originally approved to operate. This type of restriction is easiest to implement if a sys-tem of load or response monitoring is installed. Alternatively, or in addition, it may be possible to ac-cept a reduced factor of safety in the interim period until a repair is effected, i.e. a higher probability of failure during that period. This leads to a lowering of the utili-sation level in that the allowable loads are increased. Adoption of reduced design loads may also be justified by recognising that the time of exposure to environ-mental loads in the interim period may be considerably shorter than the original design life. Decision-making process Decisions about corrective actions have to be made on the basis of comparing the strength reductions Rp and Rs with their allowable values Rpa and Rsa. Fig. 7 shows an example of a flow-chart for such a decision-making process. Fig. 7: Schematic illustration of assessment process. Establish RpRp<1?No,Rp=1Estimate Up, Us,Rpa, RsaYesRp>RpaandRs>Rsa?YesReduce designloadsPossible toreduce designloads?NoYesPossible toreduce safetyfactors?NoYesNoPossible thatRs<1?Estimate RsYesNoRe-calculate Rpa,RsaRp>RpaandRs>Rsa?YesNoNoYesImpose oper-ational restriction.Repair laterEmergency repair orproceed to harbourTake necessaryprecautions.Repair laterNo further actionReduce safetyfactorsRe-calculate Rp,Rs, Rpa, RsaRp>RpaandRs>Rsa? Inspection Methods Defects and damage can only be assessed if they have been detected and measured by some sort of inspection or monitoring. Many production defects can be found by visual inspection (though that is more limited with opaque CFRP laminates than with translucent GRP). Generally for deeper defects and for most kinds of in-service damage, other non-destructive inspection (NDI) methods are needed. Comparative studies of NDI methods for marine com-posites have been performed previously. The study reported by Weitzenböck et al. (1998) considered ultra-sound, thermographic, microwave and certain acoustic methods. In the SaNDI Project new studies have been performed to include both a wider range of sandwich materials and recently developed methods, such as those based on shearography and X-ray techniques. The re-sults will be reported elsewhere, but it is relevant to mention here that some challenges remain if a fully integrated system for defect/damage inspection, assess-ment and repair is to be implemented in practice: • The potential for detecting deep defects/damage in thick sandwich structures remains limited, espe-cially for methods that can be used on board or around a ship. • It is generally not possible to detect far-side defects with one-sided inspection methods. • Thick coatings may have to be removed when ves-sels are inspected in service. •