GA MDT ON TEST FLT ALT DEVIATION.

Narrative

9/88; I WAS PIC OF AN EXPERIMENTAL MDT ACFT; CONDUCTING VNAV TESTING ON A FMS PER TYPE INSPECTION AUTHORIZATION D-8-292. IN AN EFFORT TO DETERMINE THE SUITABILITY OF THE VNAV SYS (WHICH AT THAT TIME WAS CAPABLE OF A 9 DEG DSCNT ANGLED) SEVERAL DSCNTS WERE PERFORMED ON THAT FLT. ON THE PARTICULAR DSCNT IN QUESTION; I ASKED FOR A DSCNT FROM 35000 TO 15000' AND PROGRAMMED THE VNAV TO PERFORM THE DSCNT. DURING THE DSCNT; WITH THE THROTTLES AT IDLE AND THE AIR BRAKES OUT; THE AIRPLANE WAS COMING DOWN AT ABOUT 6000 FPM OR GREATER AND REACHED VMO (370 KIAS). AT VMO (BY DESIGN); THROUGH A COMBINATION OF THE AUTOPLT/EFIS/FMS; THE VERT MODE REVERTED FROM VNAV TO AIRSPD HOLD. AT THAT POINT; THE AUTOPLT PULLED NOSE UP TO KEEP THE AIRPLANE FROM EXCEEDING VMO AND THE AIRPLANE'S ALT WENT FROM ABOUT 17500 TO ABOUT 18000'; THEN IN AIRSPD HOLD MODE TO 15000'. IN THE MEANTIME; ZME MADE THE COMMENT TO THE EFFECT THAT I HAD BEEN CLRED TO 15000'. I APOLOGIZED AND SAID THAT THE AUTOPLT ACTED UP. IN FACT; IT WAS A VNAV PROB. ZME DID NOT SUGGEST THAT ANY OTHER TFC WAS INVOLVED. AFTER THE FLT; IT WAS DECIDED BY CORP AND NAV CORP THAT THE VNAV SYS'S SOFTWARE WOULD BE CHANGED TO ELIMINATE THE POSSIBILITY OF PROGRAMMING THE SYS IN A DSCNT ANGLE THAT WOULD PERMIT THE AIRPLANE TO REACH VMO. THAT CHANGE IS DUE TO BE COMPLETED ABOUT 10/88 (CHANGE FOR MAX DSCNT ANGLE OF 9 TO 6 DEGS). CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING: RPTR CALLED IN TO RPT THAT THE FAA IS INVESTIGATING THIS INCIDENT. DURING CONVERSATION HE SAID HE HAD HIS HEAD DOWN RECORDING ACFT PERFORMANCE AND DID NOT THINK TO TELL THE ARTCC RADAR CTLR THAT THEY HAD LEVELLED OFF MOMENTARILY UNTIL THE CTLR QUESTIONED ACFT ALT.