CAVEAT EMPTOR . . . (Continued from Preceding Page)

facturer allows because of the harmonic vibration entanglements that are set up. Third, to complicate the matter further, the prop was pitched to 64 in. which is again way over the manufacturer's limit. With three strikes against it, that prop was bound to go out and

it just about took me with it. On this subject of buying props for the homebuilt, I had a long chat with Steve Wittman at Oshkosh and have subsequently talked to other well qualified, knowledgeable people. Based on what I've learned, I'd like to submit these recommendations: 1. If interested in buying a "reconditioned" prop from a shop not well-known to you, check with the FAA to find out if they are a certified prop shop. My supplier was not certified. This will not make them liable for an experimental prop that they sell you, but at least they'll use limit tables, etc., in making up your prop.

Get a yellow tag! 2. If buying a prop from a private source, check with the FAA or other facility to determine that the prop model will be suitable for your purposes. Just because it fits the bolt pattern doesn't mean a thing. 3. If buying a brand new prop, buy one that was designed for a similar application to yours. In my case, a prop for a 125-hp "Swift" might be the answer, or perhaps the "Yankee" prop. In that way, it will not be necessary to cut it down much nor will it be necessary to change the pitch much from the original. 4. If you don't know all the facts about that prop you're thinking about bolting onto your machine, stop, and don't make another move! Get the facts! Talk to the guys in the "know", look up the specs, and make certain that you aren't being eased down the primrose path. In my case, I was lucky. I'm still around to tell the tale and I fully expect to have "Tailwind" N-312S at Oshkosh in '72, but what about you? Before grabbing up that bargain, beware!!!

The failure of fixed metal propellers on homebuilts is a problem of great concern to EAA Headquarters. Several accidents of this type have occured in the past few years and have been the probable cause in some fatal accidents. In many cases the metal propeller has been a cut down, repitched one such as in the story. Often, they are damaged propellers that have been straightened before other modifications are accomplished. An added element (although not a factor in the case of Mr. Shafer's "Tailwind" accident) has been the use of these props with shaft extensions. EAA Headquarters wants to compile a report on the use (and misuse) of cut down metal propellers. We would appreciate

any and all information

members have that would apply to this particular problem area. The results will be printed in SPORT AVIATION. In addition to the advice and warnings contained in the article, EAA advises that the builder contact the original manufacturer of his metal propeller

providing the model number, the new length and pitch you plan to use, and the anticipated rpm

operating range. Ask their advice as to the feasibility of your modifications - particularly with regards to the metallurgical considerations. Secondly, consider the use of a wood propeller - the "Spit-

fire" did OJf. with wooden blades! 48 SEPTEMBER 1972

Carburetor Icing Iml Control By C. B. Shivers, Jr. (EAA 49289)

8928 Valleybrook Road Birmingham, Alabama

V^ARBURETOR

ICING,

ONE

of

the

hazards

of primary concern to pilots of light aircraft, is one of the most controversial subjects connected with the field

of aviation. My interest in this subject increased as the result of a landing, with a dead engine, at the Birmingham Municipal Airport on February 10, 1963. The aircraft involved, Piper "Super Cruiser", N-4297M, was also the first aircraft on which the Shivers carburetor ice indicator was installed. Prior to this time I had collected several informative articles relating to carburetor icing (AOPA Pilot, December, 1959 and Flying, February and December, 1960) in an effort to become better acquainted with the problems involved in an effective means of detection of ice in an aircraft engine induction system. Assuming that the information, as provided in these articles, was correct as to location of ice build-up, the only problem appeared to be the construction of a suitable detection device. The unit used provided a detection range of 36 degrees below freezing ( 32 degrees F to -4 degrees F or 0 to -20 degrees C) providing the necessary temperature detection range as suggested by these articles. Ice accumulated in this temperature range would be indicated. Initial ground test, conducted with the "Super Cruiser" during the spring of 1963, revealed that the probe wire installed adjacent to the throttle valve provided the desired indication, under simulated conditions. (The hole in which the probe was first located is provided for drilling the idle jet on the opposite side of the carburetor wall and is also the location of the Richter carburetor temperature probe and the ARP ice detector.) On November 8, 1963, after completion of the required paperwork, STC SA217SO was issued by the FAA approving the installation of the carburetor ice indicator on the "Super Cruiser". Flight test, conducted with the "Super Cruiser", soon revealed that actual ice accumulation in sufficient quantity to cause power loss did not provide an indication at the probe location. It was determined that the deflection of the fuel air mixture by the throttle valve provided a temperature consistently 15 to 40 degrees

colder than the same point (idle jet location) on the opposite wall of the carburetor, the temperature differential determined by the throttle setting. Continued test, actual and simulated, revealed that the probe provided a flat surface (the type probe presently used by the ARP ice detector) on which ice particles deflected past the

throttle valve rould be trapped by air flow to provide an erroneous indication ot' ice but did not provide an indication of ice when power loss was experienced due to actual

icing. Additional test revealed conclusively that the only point where ice accumulates, causing power loss and engine roughness, is on the throttle assembly and the probe wire was relocated on the throttle valve under

STC SA316SO. Although this information (as to ice accumulation on the throttle assembly rather than closing or

restricting the induction passage as previously depicted

in all illustrations of carburetor icing) is not presently

accepted by the FAA, despite letters and flight demonstrations including total power loss and restart demonstrations, continued test under actual conditions further substantiate the previous findings.

The vaporization process, cause of temperature reduction in the carburetor, occurs as the fuel and air mix at the point of fuel discharge, a nozzle located in the center of the venturi. As the fuel is changed from a liquid to a vapor, evaporation causes a reduction in temperature of the incoming air and the adjoining carburetor wall. The temperature drop is dependent on the throttle setting, a slight drop below outside air temperature (OAT) at idle and a severe drop at full throttle. If the incoming air contains sufficient moisture and throttle setting provides the correct temperature drop the moisture begins to

form on the supercooled surface of the carburetor wall

and discharge nozzle. Maximum condensation of moisture

on the carburetor wall or fastest rate of accumulation accompanied by a noticeable manifold pressure or rpm loss or engine roughness occurs when the temperature differential between the incoming air and the inside surfaces of the carburetor are maintained at a given tempera-

ture spread for a prolonged period of time. A simple illustration of the required temperature spread, as related to condensation, is provided by observing the condensation of moisture on the supercooled surface of a glass containing an iced beverage. It will be noted that condensation is greater during conditions of high humidity (summer months) and certain free air temperatures.

The moisture, deposited on the supercooled surface of the carburetor wall and fuel discharge nozzle, is moved toward the throttle valve by the velocity of the incoming air and the ice mass begins to build in the throttle valve

when the throttle assembly reaches freezing temperature. (Ice build-up on the throttle assembly projects into the low pressure area of the venturi and interferes with the fuel/air ratio.) The ice builds until detected by noticeable manifold pressure/rpm loss or engine roughness. Prolonged ice accumulation can result in complete

power loss occurring primarily from two causes. The first and less serious is caused by formation of ice on the

idle-jet side of the throttle valve, while operating in

cruise configuration. If a loss of manifold pressure or rpm is not noticed because of slow accumulation, the ice is detected only on closing of the throttle, if carburetor heat is not applied for a sufficient length of time to melt completely an ice accumulation prior to power reduction. (Ice accumulation on the throttle valve, idle jet side, provides a premature relationship between the idle jet opening and the throttle valve — creating suction prematurely — and supplies additional fuel from the idle jet source, causing engine failure from an overrich mixture.) Idling failure can also occur if moisture in the in-

coming air freezes as it passes the outer rim of the throttle valve, restricting the air inlet and also resulting in an overrich mixture. Corrective procedure for this type of engine failure is to advance the throttle to cruise or full open position (correct fuel/air ratio supplied by the main fuel nozzle) to restore the heat rise to the carburetor heat shroud and apply heat for a sufficient length of time to clear the ice formation, this condition indicated when the throttle can be retarded to the closed position and the engine idles normally. This type of icing was the cause of the engine failure experienced in 1963 and it was determined after landing that a throttle setting of 1600 rpm would permit engine operation; however, at the time of the landing the failure was not attributed to carburetor ice. Noticeable throttle valve icing of this type can also be expected to occur during ground idling, 600 to 1500 rpm with most aircraft, when the OAT is 40 to 60 degrees F and the relative humidity is 80 to 100 percent. Engine roughness and high magneto drop, during engine run-up, are also associated with this type of icing. If an accumulation is undetected prior to take-off, severe power loss may be experienced on take-off. If in doubt about the availability of full power, a static check should be made prior to take-off. The second and more serious type of engine failure occurs at intermediate to full power settings, when the ice mass projects into the venturi opening as described above, and has the same effect on engine operation as closing of the throttle or placing the mixture control in the idle cut off position. Under the most favorable conditions engine failure from this cause is preceded by only a slight loss of manifold pressure or rpm followed by a complete power loss resulting from a lean mixture. Although the cause of power failure is lack of fuel (improper fuel/air ratio) some fuel continues to flow from the main discharge nozzle, maintaining the temperature drop caused by vaporization, and the ice continues to build even though the engine is no longer developing power. Corrective procedure for this type of engine failure is to apply full throttle, carburetor heat, and operate the engine primer (pump rapidly) to restart the engine, also to restore the heat rise to the carburetor heat shroud. Primer operation can normally be discontinued after 30 to 60 seconds of operation or when the engine resumes normal operation without the aid of the fuel supplied by the primer. During restart procedure engine roughness may occur if all cylinders are not equipped with primer lines; however, the heat rise is sufficient to melt the ice accumulation for engine restart. This type of engine failure was first experienced on March 19, 1964, at Vero Beach, Florida while demonstrating the carburetor ice indicator to Piper Aircraft Corp. on "Cherokee", N-8500W. However, the corrective procedure for engine restart was not known at that time, necessitating a forced landing, for the second time, as a result of carburetor ice. Engine failure on this occasion occurred at 7000 ft. and all attempts to restart the engine, except primer operation, failed and a forced landing was made at the Vero Beach Airport. After ten to 15 minutes on the ground the engine started and ran normally without any modification. This demonstration and subsequent loan of Piper's Experimental "Cherokee", N-2800 to further the study of carburetor icing resulted in these conclusive findings. Use of N-2800W provided comparative study of multiple probes, in the carburetor, not allowed on standard category aircraft. This type of engine failure can be demonstrated, at any time that the OAT is 45 degrees F, on the 145-hp Continental or the 150-hp Lycoming engines. The amount of water required normally to cause total power loss, at full throttle, is three to five tablespoons. (Continued on Next Page) SPORT AVIATION 49

CARBURETOR ICING . . . (Continued from Preceding Page)

The most likely time to encounter carburetor icing, contrary to the general consensus of opinion, is during take-off, climb, and cruise configurations where heat is

seldom used unless a noticeable loss of manifold pressure or rpm occurs. Temperatures at which ice form vary with the individual aircraft because of cowling, air intake design, location of the carburetor in relation to the engine, and numerous other factors, but will be found to occur with most aircraft at 40 to 55 degrees F OAT if sufficient moisture is present in the incoming air. All aircraft, on which the indicator has been installed, have been found to accumulate ice while operating in this OAT range. The moisture content of the incoming air controls the rate of accumulation if the throttle setting and OAT provide the correct temperature differential for condensation. One of the primary sources of water, providing fast ice accumulation, is through the aircraft fuel system. Carburetor ice from this source can occur at lower OAT and at any time the water is not frozen in the fuel tanks. The improvement of fuel handling and provision of quick drains in the fuel tanks has greatly reduced carburetor icing from this cause. Use of partial carburetor heat can also cause carburetor icing when operating at lower outside air temperatures, this being the case at the time of the engine failure while test flying "Cherokee", N-8500W. It has been determined that icing is more likely to occur when the temperature, as measured at the Richter carburetor probe location, is between 0 and 12 degrees C (32 to 55 degrees F), this temperature range comprising the yellow and green arc, or caution and safe operating ranges, of the Richter carburetor temperature gauge. The application of partial heat to provide these suggested operating temperatures is more likely, under conditions of high humidity or moisture content, to cause carburetor icing at the critical point rather than prevent its formation. Further evidence of this hazard is evident in reading the article regarding engine failure as a result of carburetor icing in the September, 1965 Flying, page 104. If heat is used in the manner suggested by Richter, the reserve heat rise (full-on position) may be insufficient to clear the ice accumulation or addition of the heat reserve may increase the rate of accumulation. With the information provided during developmental testing the suggested use of carburetor heat would be full heat if engine roughness or a loss of manifold pressure or rpm occurs, returning the heat control to the off position when the manifold pressure or rpm is regained. Carburetor heat should also be applied at least one to two minutes prior to power reduction and for a sufficient length of time to melt any undetected ice accumulation. If no engine roughness occurs, that would indicate ingestion of water by the engine, the carburetor heat can be placed in the off position if desired. After power reduction very little heat is provided from the carburetor heat shroud and is inadequate to melt an accumulation present prior to power reduction. Utilizing the Shivers ice indicator the pilot is warned of ice accumulation prior to the time a noticeable loss of manifold pressure or rpm occurs (normally five to 15 minutes depending on the rate of accumulation), the presence of an ice accumulation of minute thickness providing an indication. Ice formation is indicated by movement of the indicating needle into the red arc as the ice formation begins to bridge the gap between the probe electrode and the throttle valve. When the indicating needle moves into the extreme left edge of the red arc the ice mass has increased in size and will begin to affect engine operation if it is not cleared. Prolonged ice accumulation, allowing the indicating needle to 50 SEPTEMBER 1972

(Photo Left) The redesigned gauge and throttle assembly which comprise the complete carburetor-ice-indicator installation kit. (Photo Right) Marvel-Schebler MA-4-5 carburetor reflects the installed modified throttle assembly. This carburetor is commonly used on the Cessna 182, Mooney "Statesman", and Piper PA-28-180.

The black area on the throttle plate indicates the ice build-up area on the throttle fly. The dotted lines indicating the fly in closed position show how the same build-up of ice will close off the idle jets.

remain in the extreme left edge of the red arc for an extended length of time, can result in complete power failure as described above, necessitating restart procedure. Application of full carburetor heat and clearing of the ice formation will be indicated when the indicator needle returns to the green arc. Early detection of ice can prevent damage to the engine as a result of prolonged operation with a lean mixture (slow accumulation may prevent detection

through roughness or loss of manifold pressure or rpm). Formation of ice on the throttle valve, cause of the lean mixture, can occur when the mixture control is in the full rich position. A slight roughness from this cause may not be noticed by the pilot. Minimum use of carburetor heat will save fuel and increase range. In addition, for maximum power, the densest or coldest air is desirable, and the Shivers ice indicator allows the maximum use of cold and filtered air (except when heat is actually needed to combat icing).

Unlike the carburetor temperature gauges, the car-

buretor ice indicator is dependent on neither temperature nor humidity, only the actual presence of an ice formation for an indication. £)