O-320H Cam Spalling -- Accident or Suicide?
by Bill Marvel and Bill Scott
Readers of our major work on this web site, The Rest of the Story, are aware of our findings of very low oil volume provided to the rocker boxes of Lycoming engines with mushroom style hydraulic lifters. It is this low oil flow that we have related to excess exhaust valve guide wear in several engines we have been testing. This low oil flow condition results from the fact that the mushroom style lifter was originally designed for automotive use in an engine configuration which did not require the lifter to pump oil at all. Accordingly, it has no means for doing so.
When Lycoming subsequently modified this lifter for use in their aircraft engine, they did not provide any unrestricted, dedicated oil flow path to the rocker boxes. As a result, oil flow to the rocker boxes is minimal to all cylinders, and particularly to those on the copilot side of the engine, which experience most of the valve and guide failures. The reason for this side-to-side oil flow variation is explained in detail in the above-mentioned article.
Our goal is to attain a full understanding of the nature of this problem and to find a future solution for it. In the process of our investigation, however, we stumbled upon something of great interest, although not directly related to what we are trying to achieve. This discovery has to do with a possible cause of, or certainly a likely contributor to, the valve lifter and cam distress problems that appeared in the early O-320H engines.
As a brief background, during our research we learned that Lycoming's version of the mushroom style hydraulic lifter, which is used in almost all of their engines, has no provision for pumping or transferring substantial amounts of oil to the rocker box. In sharp contrast, we pointed out a major design difference in Continental's version of this same component which does provide for a clear oil flow path to the rocker boxes of Continental engines.
Later, our testing revealed that the barrel type of lifter, now used extensively by Continental but much less by Lycoming, sends a substantially greater volume of oil to the rocker boxes than does the mushroom type of lifter. The reason for this is that the barrel lifter was actually designed for use in overhead valve engines, the configuration employed by both Lycoming and Continental, whereas the mushroom lifter was not. Specifically, the barrel lifter can perform two necessary functions continuously and simultaneously -- maintaining near zero valve train lash and sending a large volume of oil through the pushrod to the rocker box. In contrast, the mushroom type of lifter can only maintain valve train lash. It does not have the ability to pump or transfer oil anywhere.
Because Lycoming's design of their version of the mushroom style lifter lead directly to the low rocker box oil flow we have found, we wondered if any similar design aspect could have resulted in the early problems with O-320H engines. Many of you will recall the serious cam and lifter spalling occurrences in the early life of the O-320H when it came out in the Cessna 172 in 1977.
The O-320H engine was equipped with the barrel type of lifter and in the field experienced immediate problems with cam lobe and lifter spalling. Lycoming changed lifter designs several times and eventually went to both a different lifter and cam shaft before the problem was solved. Because the most important function of the lifter is to maintain a near zero valve train lash, and because loss of lash control will cause the type of cam and tappet damage that the O-320H encountered, we had a simple question. Was there anything in the design of the early O-320H lifters that could be identified as possible explanations for the type of damage these engines were sustaining?
To refresh your memory, look for a moment at figure 7 in our major work, The Rest of the Story. This is a line drawing of a typical automotive barrel style hydraulic lifter, which is functionally identical to what both Lycoming and Continental use today. It is important to note in this depiction that pressurized oil from the main engine gallery is available to the lifter 100% of the time via the full registration annulus (4) and that this pressurized oil provides the mechanism for accomplishing the lifter's two functions:
1. Maintaining a near zero valve train lash by flowing through the check valve (8), and
2. Sending oil continually through the pushrod via the pushrod socket oil passage (9) for both lubrication and cooling of the rocker box components.
(If for any reason oil pressure to the lifter is lost or greatly reduced, both of these functions will be compromised to some extent)
It is with the above two functions in mind that we began to research Lycoming's service publications regarding the sequence of replaced lifter part numbers so that we could attempt to acquire some or all of them. The first of these was Service Bulletin 424 dated March 24, 1978. This bulletin referenced a kit composed of either P/N 16168 or 16585 lifters. The kit consisted of 8 of one lifter P/N or the other and required replacement of all existing lifters due to spalling.
A year later, on March 17, 1979, Service Bulletin 435 was issued. This S.B. provided for the replacement of the 16168 or 16585 lifters changed a year earlier with a new part number, 16586. However, only the exhaust lifters were called out for replacement and accordingly, only 4 lifters were provided in this kit.
When problems continued, Lycoming published Service Instruction 1402 on September 5, 1980, which incorporated larger diameter lifters, P/N 16812. This is the lifter currently in use in the engine. Cam and lifter spalling is now rarely a problem in the O-320H series engines.
In our initial reading these publications, something very curious jumped out at us. Service Bulletin 435 contains the following:
NOTE The exhaust hydraulic lifter P/N LW-16585 or P/N LW-16168, ....... cannot be reused and must be destroyed.
Must be destroyed? Really? Why must the superseded parts be destroyed? We aren't told to destroy replaced cams or pistons or spark plugs. What was the reason for this requirement? Our curiosity escalated.
It took a number of phone calls and some digging, but we were able to locate a 16585 lifter referenced in S.B. 424 and a 16586 lifter referenced in S.B. 435. Of course, we were easily able to obtain the current lifter, P/N 16812 but were never able to find an example of the 16168 part mentioned in S.B. 424.
On receipt of the two superseded lifters, 16585 and 16586, a quick glance immediately revealed a surprising design characteristic that we will show you in a moment.
First, however, look at the following:
This depicts the three lifters in order of replacement from left to right -- 16585, 16586 and 16812. They are shown in the same orientation as figure 7, with the pushrod socket at the top and the face (which rides on the cam lobe) at the bottom. In this photo the left and center lifters appear virtually identical to each other. The newest lifter on the right, P/N 16812, is larger in diameter and has a slightly different annulus shape and location. The interior components of all three, upon detailed inspection, appear to us to be identical.
However, rotating all three lifters 180 degrees about their vertical axis reveals something that everyone reading this should immediately recognize as a potential problem. Note the following:
In this photograph, the two superseded lifters are shown to have been manufactured with a machined flat along a portion of their outer surface. Careful observation of this flat reveals that it connects with the annulus of the tappet body on one end, and extends to near the tappet body face at the other end. It is clear what this flat is intended to do -- bleed pressurized oil away from the lifter in order to spray it onto the cam lobes for added lubrication. But doing so creates the exact problem we mentioned above, which is that any reduction of oil pressure going into the lifter annulus compromises its ability to perform the two functions demanded of it -- lash control and routing oil through the pushrod to the rocker boxes. And yet both of the superseded lifters contain machined flats which do exactly that.
There is an irony here. In The Rest of the Story, we highlighted our one-sentence summary of the core problem we found with the mushroom style lifter:
When Lycoming adapted the automotive mushroom style hydraulic tappet assembly to their engine, they failed to provide a dedicated, unrestricted oil flow path to the rocker boxes that could carry sufficient oil for adequate cooling as well as for lubrication.
The irony is that in the early, superseded O-320H lifters, Lycoming did create a dedicated, unrestricted oil flow path when they should not have. Unfortunately, this path allowed pressurized oil to escape from the entry point to the lifter, thus compromising its ability to function.
Are we saying conclusively that this is the whole story with the O-320H? Not at all. In addition to lifter changes, the company eventually also installed a new cam shaft with wider lobes. Clearly, the combination of wider cam lobes and a larger diameter tappet body distribute valve opening loads over a larger surface area and this may well have played a role in minimizing spalling incidents. Additionally, we have no way of knowing what hardness or other metallurgical characteristics may have been changed along the way. We find it of interest, however, that hydraulic lifter design characteristics seem to be a common point which ties together our discoveries and the early O-320H problems.