3003 W CASINO RD BLDG 45-03 2019-01-28

Geminus F aject ccetractS o.EVE-ib-2920 Terhnelvgy Development I CM Project No.160731 5 Summary/Conclusions 5.1 General Airflow& %LEL CFD Modeling91 45-03 Hangar Comparison of the CFD computed air flow velocities around the 777-300ER aircraft in the 45- 03 painthangar with data from the TAB shows that the CFD airflow model agreed very well with the information available from the TAB report to within the accuracy of the measurements. A maximum velocity error in the CFD analysis of 66%can be estimated by analysis of the TAB and CFD data at the crown and fuselage. That is,the actual velocity in the hangar could be up to—3x slower than predicted by the CFD(see Section 3.4.1).. The subsequent CFD model predictions of%LEL distributions during normal paint mode clear- coat application operations in the 45-03 paint hangar are consistent with historical measurements in 45-01 (which has a very similar interior geometry)-i.e.,very low levels(<<1%LEL)away from the aircraft(e.g.,wing tips)and very low levels(<l%LEL)in the floor exhaust ducts. The reported 8 hour time-weighted average%LEL measurements in the 45-01 indicated 0%LEL at all locations which is less than the low, but non zero, values predicted by the CFD model— i.e., steady-state values in the range of 0.05 to 0.5 %LEL in similar areas as the measurements. Peak unsteady %LEL values can be derived from the CFD model computations by multiplying by a factor of —1.6,thus bringing the estimated peak values close to those measured during brief moments in the 45-01. Mixing plume analysis shows that dilution of volatiles in a jet is proportional to the velocity of the jet Using the 3x uncertainty factor on the velocity,the!/CLEL could be estimated to be up to 3x higher than the CFD predicts. The factor of 1.6 due to turbulence could be applied to this as well,resulting in an overall%LEL uncertainty of--5x(see Section 3.3.3 and Section 3.4.2). 5.2 %LEL Predictions br,New 45-03 Hangar Configuration. Nominal paint mode produced%LEL levels approximately the same as those predicted in paint mode for the pre-modification 45-03 and are consistent with those measured in the 45.01 (see previous references and sections). Reduction in airflow by 25%to the 75%level did not show a non-linear increase in local or overall %LEL level as the levels were elevated by an expected proportional amount with the volumes of%LEL growing as expected. Removing cowl painting operations significantly reduces the%LEL levels in the hangar volume,particularly in the low bay areas,and may be a method to further ensure low%LEL levels during painting operations. The fuselage and 75%air flow paint modes produce comparable%LEL levels to the nominal paint mode. It was shown that the overspray from the engine cowl painting had a large effect on the amount of overspray (%LEL) present in the low bay of the hangar due to the lack of floor exhaust in close proximity. It was also shown that the CFD model predicts <1%LEL levels throughout the paint hangar (Case 21Scenario 2) in the event of a full shut down of the air supply/exhaust system after approximately 120 seconds from the shutdown event. It also predicts that%LEL concentrations during the shutdown process do not increase in any local or overall value during the air shutdown process under the assumption of rapid painting stoppage_ This is a result of the very long air supply and exhaust spool down times which produce continued dilution of any residual overspray long after painting stoppage. Additional modifications are made to the hangar air supply system in Case 3 Scenarios 2;4b, and Sd:the engine ceiling diffuser is moved 11' outboard and two additional diffusers are placed over the outboard tip of each wing. The results from these CFD simulations show that these