A CARJ FO RPTED AN ALTDEV AND TRACK HDG DEV APPROX 120 NM NW OF BKW VOR.

Narrative

WE LEFT CVG VIA THE TODUB 2 DEP. AFTER SEVERAL VECTORS; ZID CLRED US DIRECT BKW VOR (BECKLEY). THE CAPT (PF) ELECTED TO HAND FLY THE ACFT TO ALT. ATC ISSUED SEVERAL INTERMEDIATE ALTS; WHICH WERE NEVER REACHED BEFORE A HIGHER ALT WAS ASSIGNED. SOMEWHERE AROUND FL120; THE PF TURNED ON THE FLT DIRECTOR. THE ACFT WAS PERFORMING WELL (300 KIAS-320 KIAS AND 1500 FPM) IN THE CLB; MUCH BETTER THAN NORMAL (290 KIAS AND 500-1000 FPM); MOSTLY DUE TO BEING LIGHT AND THE COLD WX. THE FLT DIRECTOR WAS SET TO A 500 FPM CLB. THE FLT DIRECTOR BARS GRADUALLY SANK TO 5 - 7 1/2 DEGS BELOW THE HORIZON; WHILE THE PLANE WAS AT 2 - 2 1/2 DEGS ABOVE THE HORIZON. THE PF GRADUALLY LET THE PLANE DRIFT SEVERAL DEGS L OF COURSE. THE FLT DIRECTORS (SET TO 'NAV' MODEL) BEGAN A GRADUAL BANK TO THE R UNTIL THEY WERE 25-30 DEGS TO THE R; STILL SHOWING 7 1/2 DEG PITCH. THE CONFLICTING MESSAGES (FLT DIRECTOR BURIED AND TANKED VERSUS THE ACFT ON COURSE [MOSTLY]; AND PITCHED UP AND CLBING QUITE WELL) LED TO CONFUSION OF THE PF; AND A RELUCTANCE TO ENGAGE THE AUTOPLT UNTIL THE PERCEIVED 'FLAIR' WAS STRAIGHTENED OUT. THE PF BEGAN CORRECTING COURSE BACK TO THE R; BUT DID SO EXCESSIVELY; ABOUT 30 DEGS TOO MUCH. THE FLT DIRECTOR COMPUTER LAGGED AND DIDN'T REFLECT THE COURSE CORRECTION; AS THE PF THOUGHT IT SHOULD. ABOUT THAT TIME; FL270 (HIGHEST ASSIGNED) WAS REACHED; AND THE PF BEGAN ANOTHER COURSE CORRECTION BACK L. THE ACFT DRIFTED UP TO 27300 FT. AS SOON AS THE ACFT LEFT 27000 FT; A TCASII 'TFC; TFC' MESSAGE SOUNDED; THE DISPLAY SHOWED AN ACFT AT OUR 3 O'CLOCK POS AND 5-7 MI 700 FT ABOVE US XING OVERHEAD. ATC ASKED US TO VERIFY ALT; WHICH WE DID. IN THE MIDST OF THIS CONFUSION; THE ACFT WAS ALLOWED TO DRIFT 60 DEGS L; ENDING UP 30 DEGS L OF COURSE NOW. PNF WAS STILL LOOKING FOR TFC XING OVERHEAD. THE AUTOPLT WAS ENGAGED AT FL270; AND THE FMS RESET TO INDICATE BACK DIRECT BKW VOR. ATC QUERIED US AS TO OUR NAV AND WE RESPONDED WE WERE PRESENTLY DIRECT BECKLEY. THIS PROB RESULTED FROM A MISUNDERSTANDING/LACK OF UNDERSTANDING ABOUT THE FLT DIRECTOR COMPUTER AND ITS REASONING; CALCULATIONS; AND INDICATIONS. BETTER CRM COULD HAVE HELPED TO PREVENT THIS PROB (IE; THE PNF SHOULD HAVE BEEN MANIPULATING THE FLT CTL PANEL CTLING THE FLT DIRECTOR). EVEN IF A PLT IS SMOOTH ON THE CTLS; IT IS DIFFICULT TO CTL THIS JET THAT IS ENTIRELY CTLED BY HYDRAULICS WITH ARTIFICIAL FORCE FEEDBACK. AT HIGH SPDS AND HIGH ALTS; PLTS WOULD BE BETTER OFF AND HAVE THEIR WORKLOAD REDUCED; BY USING THE AUTOPLT AND LEAVE 'PROFICIENCY FLYING' TO THE SIMULATOR.