Narrative:

Third leg of the day for a night landing. First officer (first officer) was flying ILS to runway xy which does not have a displaced threshold. It was a clear night and we could see the runway and PAPI prior to glideslope capture. As we intercepted the localizer to runway xy; the first officer's flight director went away. I still had both HUD and pfd flight directors and told him so. I asked him if he was ok to fly the raw data and he replied yes. His reply was quite affirmative. I didn't detect any hesitation. I told him I would back him up. Later; when we discussed the approach; he told me that he had been confident that he would be fine flying with just the raw data. It turns out that he struggled with maintaining glideslope. We remained on centerline the whole way to touchdown. But about three times got a dot low. Each time he corrected to get back on glideslope; but only got there the first time. As I looked out at the PAPI; they never showed us low when we actually were according to the ILS. I assumed he was looking at the PAPI and runway more than the raw data. It would explain why he was going a little low and not catching it right away until he referenced the raw data. It was also my perception at the time that this is why he was so steady on lateral course. This was a bad assumption on my part because the PAPI eventually showed us low and he did not correct as I would have expected. We are now on short final; and the PAPI went to a dot low; then as it hit 1.5 dots low we got the 'glideslope' callout which should have come at greater than a dot low. The ILS was also showing us now 1.5 dots low. I called out that he was 1.5 dots low. He flattened out the vertical speed until we were about a dot low and we landed about 500-700 feet down the runway. I had told the first officer that I would back him up; yet said nothing during the approach until we got the 'glideslope' callout. I have always found raw data to give me the first hint of deviating and use it to this day even with a HUD. My expectation was that the first officer would have no trouble VMC with the runway in sight using raw data to maintain a 3 degree glideslope. I was wrong. He admitted the next day that he had struggled more than he expected with just the raw data. Contributing to the deviation: 1. I allowed my comfort level with going low to be excessive since we were landing at a runway preceded by open water. 2. I allowed his total hours in the airplane to give me a false sense of 'he knows what he is doing.' the next day; when we discussed it; his perception of what happened was that the airplane experienced a series of 'sinkers' to which he responded with adding power. This was not what was actually going on. He was establishing something greater than a 3 degree glideslope. We were not getting slow and there was no shearing effect. By adding power he was getting some pitch up from vector thrust; but as soon as he reduced the power; we were back to the slightly excessive glideslope. I didn't recognize it at the time; but looking back and after discussing it with him; this is precisely what was happening. His visual perception of what the glideslope should be was off similar to the effect of vertigo on attitude. The lack of ground visual over the dark water was probably the reason. I asked the first officer if he had ever flown raw data without a flight director in either the sim or airplane before. He couldn't remember for sure; but didn't think so. This might be something worth considering in training for both pilots. What will I do in the future?1. Stop assuming. From now on in my first preflight brief of a sequence I will explain that it is too risky to make any assumptions for either of us and we will verbally address all concerns. 2. With the loss of a first officer flight director VMC I will either take the airplane; or actually back up with deviation callouts. IMC I will take the airplane.3. My comfort level with being low will not exceed a half dot regardless of the terrain. In terrain airports; itwill be 1/4 dot. There is just no need to accept more.

Related reports:

Google
 

Original NASA ASRS Text

Title: B737-800 flight crew describes a night VMC ILS flown by the First Officer with his Flight Director inoperative. The First Officer has difficulty maintaining a proper glideslope using raw data and the PAPI. On short final a glideslope warning is heard and 1.5 dots low is noted by the Captain. A thrust adjustment is made to land in the first 500-700 feet of the runway.

Narrative: Third leg of the day for a night landing. First Officer (FO) was flying ILS to Runway XY which does not have a displaced threshold. It was a clear night and we could see the runway and PAPI prior to glideslope capture. As we intercepted the localizer to Runway XY; the FO's flight director went away. I still had both HUD and PFD flight directors and told him so. I asked him if he was ok to fly the raw data and he replied yes. His reply was quite affirmative. I didn't detect any hesitation. I told him I would back him up. Later; when we discussed the approach; he told me that he had been confident that he would be fine flying with just the raw data. It turns out that he struggled with maintaining glideslope. We remained on centerline the whole way to touchdown. But about three times got a dot low. Each time he corrected to get back on glideslope; but only got there the first time. As I looked out at the PAPI; they never showed us low when we actually were according to the ILS. I assumed he was looking at the PAPI and runway more than the raw data. It would explain why he was going a little low and not catching it right away until he referenced the raw data. It was also my perception at the time that this is why he was so steady on lateral course. This was a bad assumption on my part because the PAPI eventually showed us low and he did not correct as I would have expected. We are now on short final; and the PAPI went to a dot low; then as it hit 1.5 dots low we got the 'glideslope' callout which should have come at greater than a dot low. The ILS was also showing us now 1.5 dots low. I called out that he was 1.5 dots low. He flattened out the vertical speed until we were about a dot low and we landed about 500-700 feet down the runway. I had told the FO that I would back him up; yet said nothing during the approach until we got the 'glideslope' callout. I have always found raw data to give me the first hint of deviating and use it to this day even with a HUD. My expectation was that the FO would have no trouble VMC with the runway in sight using raw data to maintain a 3 degree glideslope. I was wrong. He admitted the next day that he had struggled more than he expected with just the raw data. Contributing to the deviation: 1. I allowed my comfort level with going low to be excessive since we were landing at a runway preceded by open water. 2. I allowed his total hours in the airplane to give me a false sense of 'he knows what he is doing.' The next day; when we discussed it; his perception of what happened was that the airplane experienced a series of 'sinkers' to which he responded with adding power. This was not what was actually going on. He was establishing something greater than a 3 degree glideslope. We were not getting slow and there was no shearing effect. By adding power he was getting some pitch up from vector thrust; but as soon as he reduced the power; we were back to the slightly excessive glideslope. I didn't recognize it at the time; but looking back and after discussing it with him; this is precisely what was happening. His visual perception of what the glideslope should be was off similar to the effect of vertigo on attitude. The lack of ground visual over the dark water was probably the reason. I asked the FO if he had ever flown raw data without a flight director in either the sim or airplane before. He couldn't remember for sure; but didn't think so. This might be something worth considering in training for both pilots. What will I do in the future?1. Stop assuming. From now on in my first preflight brief of a sequence I will explain that it is too risky to make any assumptions for either of us and we will verbally address all concerns. 2. With the loss of a FO flight director VMC I will either take the airplane; or actually back up with deviation callouts. IMC I will take the airplane.3. My comfort level with being low will not exceed a half dot regardless of the terrain. In terrain airports; itwill be 1/4 dot. There is just no need to accept more.

Data retrieved from NASA's ASRS site and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.