Heads Up – K Model Head Quandary?

Production K Models

Re: Heads Up – K Model Head Quandary?

Postby EKHKHK56 » Sun Apr 30, 2017 7:21 pm

To associate the K with failure of any type as a design and actual motorcycle to me is pure baloney. Monday morning quarterbacks looking back 55 years. The Model K was the most beautiful motorcycle ever designed in many enthusiasts opinion, an all new design incorporating the best styling cues from H-Ds other models with new modern suspension, unit case and large brakes etc. :)
Ask anyone back in 1952 about this beauty. It was instantly loved by many around the world, motorcycle lovers took note! The K has evolved into the most successful model motorcycle ever in the H-D line up. My OP 52K went 20 years and almost 20,000 miles without a transmission problem or other major problem. Than the shifter spring broke inside. Sure any idiot can pop the tranny and cases drag racing or other abuse. Just like all these Hemi powered cars and Jeeps can rip the 904 w overdrive transmissions out in one abusive test drive. Operator error causes most failures attributed to bad designs. As for using the same part number for new parts, it makes perfect sense if the part is the same one with only minor modifications or improvements. Obviously the head numbers on heads are a bit off yes. As we know H-D usually added the -A, -B etc to identify them as a newer version of same part inless a major change occured than they would change the year suffix when that happened. Also they might change one number on the end for similar parts. Except in cases where production changes out paced the book writers or the foundry typesetters, and and an existing number was floated. There probably should have been -52, -52A, -54 & -54A numbers for suffixes. The Spare Part book does list the later heads as different numbers with last number one higher than -52# and with the -54 suffix for 54-56 KHK. It lists the -52 heads for 52-56 K, KK, & KH. The -59 Book. I think the 54KH had the newer style head but weak raised numbers, contradicting the book again. So the foundry or whoever presented the material to the foundry made that error in number 54 upgrade. Or we've just never seen a -54 head :mrgreen: Early heads have stamped in numbers, the 54ish ones have weak raised numbers and the 55 56 have tall sharp raised numbers. Looking at the 32 heads I have. There are KK heads like Patrick mentioned. Deep relieved but not wide on intake valve. No polishing. Early heads have smaller chambers with shallower relief than newer ones definitely. I think it is .040" less on a K in the intake valve head area. More fun K "what's really going on thoughts.." Cheers folks and RIP to my Buddy Joe Leonard who we lost Thursday afternoon. He was a close friend and great fellow. I miss the 4-5 emails a day etc. :| He was so interesting and friendly besides the greatest racer ever. No one has matched Joe's Multiple Consecutive National Championships on Motorcycles and in Cars. Erik K
User avatar
EKHKHK56
 
Posts: 920
Joined: Fri Apr 18, 2014 5:20 am
Location: Fairbanks, Alaska USA

Re: Heads Up – K Model Head Quandary?

Postby sportsterpaul » Mon May 01, 2017 7:03 am

Wow, 32 heads, that is fantastic. You are a national treasure, lets hope Russia does not try to take Alaska back. When I look at the original post again, it may not be that the E with a circle around it means the letter revision of the casting part number. If its a rev letter, where are A, B, C, and D? The letter E could just mean what foundry made it, or what pattern or some other code. The circle and the distance to the base part number implies it in not a revision letter. In that case it could mean that Harley never did change the casting number from 16684-52, but they did change the foundry. The "E" foundry made some of the old heads, and then Harley sent the new drawing, and yeah, maybe with the same number, and had the foundry make patterns for that. I think we agree when you say "So the foundry or whoever presented the material to the foundry made that error in number 54 upgrade."

It was a simple error and not a tragic one as long as the finished part numbers that go in the parts book and on the BOM are different. Is it agreed that the finished part numbers in the parts book correspond to the high- and low-window heads? As to the valve pockets, it is perfectly reasonable that the same casting number might have different machining. Machining is a second op after casting, and would not change the casting number, but would likely change the finished part number that goes in the parts book.

Now this would be a lot more reasonable. When the engineers do the layout drawing to make sure everything fits, they don't use the casting number drawing, but the finished part drawing. So its kind of a "buried problem" having different casting parts with the same number. As long as it is an internal Harley issue, and not something that affects service or replacement parts, I could see them just ignoring it.
sportsterpaul
 
Posts: 240
Joined: Mon Jul 06, 2015 2:17 am
Location: Sun City Center, Florida

Re: Heads Up – K Model Head Quandary?

Postby wz507 » Mon May 01, 2017 11:25 am

EKHKHK56 wrote:Also they might change one number on the end for similar parts. Except in cases where production changes out paced the book writers or the foundry typesetters, and and an existing number was floated. There probably should have been -52, -52A, -54 & -54A numbers for suffixes. The Spare Part book does list the later heads as different numbers with last number one higher than -52# and with the -54 suffix for 54-56 KHK. It lists the -52 heads for 52-56 K, KK, & KH. The -59 Book. I think the 54KH had the newer style head but weak raised numbers, contradicting the book again. So the foundry or whoever presented the material to the foundry made that error in number 54 upgrade. Or we've just never seen a -54 head :mrgreen: Early heads have stamped in numbers, the 54ish ones have weak raised numbers and the 55 56 have tall sharp raised numbers. Looking at the 32 heads I have. There are KK heads like Patrick mentioned. Deep relieved but not wide on intake valve. No polishing. Early heads have smaller chambers with shallower relief than newer ones definitely. I think it is .040" less on a K in the intake valve head area.

I am not disagreeing with any of the above since I certainly have no answer to any of this. As a matter of fact I like the above explanation, as it follows my initial thought process and I think might be correct. However, that was when my knowledge of the subject consisted of only the 4 heads in my possession, and since then I’ve seen exceptions. Specifically, using the guidance above, the “circle E” head shown below with the “positive” tall sharp casting numbers would be considered a ’54 to ’56 head, yet in fact it has all the major features of the earliest head with the negative “underscore 1”casting number, i.e., tall window and coarse discontinuous fins. So clearly this head although of “late” casting type is early in all other aspects. Still makes me wonder if all K heads from early to late could be correct for all years of K models?

Another thing I noted relative to the early head, with so-called "negative casting" number and having no combustion chamber machining was that it appears to allow the use of a large 1.800" in IN valve with no issue. To make this assessment an early head was fit with a thin layer of clay around the valve circumference, the head in turn fit to a cylinder and the IN valve pushed into the clay to show the registration of the valve in the combustion chamber. As the last image below shows, there is generous clearance around the circumference of both valve heads. Perhaps a pass with a burr might be in order to open the chamber ever so slightly in the lower IN plain between the IN and EX valves, as it appear that once upon a time a valve might already have floated up into the head in this region.

As always eager to hear others thoughts on correct heads for a '56 KHK. Thank you.

lr circle e early combustion chamber.jpg
lr circle e early combustion chamber.jpg (153.42 KiB) Viewed 11787 times


lr circle E head.jpg
lr circle E head.jpg (52.44 KiB) Viewed 11787 times


lr early clay head.jpg
lr early clay head.jpg (187.49 KiB) Viewed 11787 times
User avatar
wz507
 
Posts: 356
Joined: Fri Dec 11, 2015 9:13 pm

Re: Heads Up – K Model Head Quandary?

Postby sportsterpaul » Mon May 01, 2017 12:47 pm

The 1954 in the gallery section
http://www.harleykmodel.com/gallery/sur ... index.html
has the smaller window-- like the head on the right in your original post. There are a couple 53 bikes in the gallery where you can see the fin window and both of them and my 52 has the bigger window.

Its a pretty good bet that your 56 takes the smaller window, and that is the -54 finished part. I assume the more metal at the bottom was trying to prevent the cracking that Patrick figured out. I also wonder if the change in the casting is not as large as I first thought. Changing the fin on top could be a simple glue-up to the wooden pattern. The size of the window is set by a core that goes in the mold. (Patrick, does this make sense?) So maybe they left the casting number the same, and since they didn't put rev letters on the part, the base number didn't change.

Patrick also explained how the patterns were made by the foundry, not at the factory. So the fact that the two heads on the left on the original post are a tiny bit different makes sense. If Harley changed foundries, I could see the first one not giving up their patterns, so the "E" foundry made a new pattern to the older design and ran those, maybe in 1953. Then in 1954 the change came in, and the "E" foundry made or modified the core and changed the pattern to make the small-window head with the different fin on top.
sportsterpaul
 
Posts: 240
Joined: Mon Jul 06, 2015 2:17 am
Location: Sun City Center, Florida

Re: Heads Up – K Model Head Quandary?

Postby wz507 » Mon May 01, 2017 12:49 pm

With the 3 different heads in my possession was curious to know a bit more about the combustion chamber thickness in the valve area, the depth of the combustion chamber, the maximum valve lift allowed, and the fit of the valves in the chamber. In the data presented below the heads are numbered 1 through 3 and correspond to their appearance in the original pictures where the left head (oldest?) is #1, the middle head (intermediate?) is #2 and the right head (newest?) is #3.

To address the head thickness it was reasoned that if one knows the total height of the head above the combustion chamber and the chamber height then the head thickness is also known. To make this assessment it was assumed that the machined head bolt spots would conservatively represent the total head height since the machined spots are slightly below the top of the head casting. To make this measurement each head was positioned on its sealing surface and depth measurements were made at each of the 5 machined head bolt holes surrounding the valve area (see image below) to determine total combustion chamber height. The extreme values in head height measured across all 3 heads were 0.918” and 0.942” with the average of the 5 values for each head shown in the table below. Based on the average values measured one might assume that the factory specification for the machining operation might have been 15/16” (0.9375).

The heads were inverted and combustion chamber depth measured at points A through D as shown in the image below. Recall that the valves are tipped in 2 plains, i.e., both lengthwise in the chamber and crossways in the chamber. Thus the crossways measurements made along the constructed line will be deepest at point B for the IN valve and point D for the EX valve. For each head the extreme measurement at point B was subtracted from the total head height to provide the minimum combustion chamber thickness of the 3 heads, which ranged from ~0.425 to ~0.465”. So lots of meat above the combustion chamber.

lr measurement locales.jpg
lr measurement locales.jpg (105.2 KiB) Viewed 11787 times

To measure the maximum valve lift potential of each head, a 1.800” IN and 1.572” EX valve were installed in the rear cylinder, weak springs fitted, and valve lift determined by moving the valve through its maximum range of motion while measuring with a dial indicator. When seated, the highest point on both IN and EX valves were near, but slightly below the top plain of the cylinder deck.

lr dial indicator.jpg
lr dial indicator.jpg (119.93 KiB) Viewed 11787 times

Of course ambiguity exists in making the measurements described here since they are impacted by a variety of variables including the depth of the valve seat (the depth the valve has been sunk into the cylinder with repetitive valve jobs), the margin remaining on the valve, etc. Although the foregoing factors affect the maximum lift observed, they do not however affect the difference in maximum lift observed between the various heads, which remains constant. The measured maximum lift results are included in the table.

Table data for K model head.jpg
Table data for K model head.jpg (26.19 KiB) Viewed 11787 times

The results suggest that any of the K heads from the earliest to the latest would be suitable for any K model application since they all have essentially the same head height, similar combustion chamber thickness, and can accommodate the large IN valve (see image in 2nd post above) and the 0.375” KHK/KK cam lift. It is my opinion that with machining any early head could be readily modified into any later head configuration. This is likely old info to many, but new to me as I was operating under the misconception that the early heads were for only small valves and low lift cams, when in fact all K heads appear to be created equal. But, the caveat......... the above comments are predicated on the examples I've seen and measured and very likely there are heads I've not seen, thus "you don't know what you don't know". But I'm always eager to hear more, so please help me better understand what I don't know. Thank you.
User avatar
wz507
 
Posts: 356
Joined: Fri Dec 11, 2015 9:13 pm

Re: Heads Up – K Model Head Quandary?

Postby wz507 » Mon May 01, 2017 1:52 pm

Paul, thanks for your comments.

sportsterpaul wrote:The 1954 in the gallery section
http://www.harleykmodel.com/gallery/sur ... index.html
has the smaller window-- like the head on the right in your original post. There are a couple 53 bikes in the gallery where you can see the fin window and both of them and my 52 has the bigger window.

But I also see in the gallery a 1954 KHK (Chuck Sherman) and a 1955 KHRM with tall windows and I'm sure if we kept surveying sooner or later we'd find a bike with one of each. ;)

sportsterpaul wrote:Its a pretty good bet that your 56 takes the smaller window, and that is the -54 finished part. I assume the more metal at the bottom was trying to prevent the cracking that Patrick figured out. I also wonder if the change in the casting is not as large as I first thought. Changing the fin on top could be a simple glue-up to the wooden pattern. The size of the window is set by a core that goes in the mold. (Patrick, does this make sense?) So maybe they left the casting number the same, and since they didn't put rev letters on the part, the base number didn't change.

I agree that a '56 likely does take the short window head with long continuous runs of fins on the rear head surrounding the sparkplug. And then again there is the Circle E casting having a short window, but early discontinuous fins, and a fully machined combustion chamber ala KHK. God only knows, and even he may not fully understand K model heads.

Circle E with short window and coarse fin.jpg
Circle E with short window and coarse fin.jpg (38.12 KiB) Viewed 11774 times


Circle E - short window discontinuous fin fully machined CC.jpg
Circle E - short window discontinuous fin fully machined CC.jpg (49.37 KiB) Viewed 11774 times

Regarding cracking, although extra material in the vicinity of the window might address a strength issue, all the heads from early to late are essentially the same thickness (see post above). Furthermore, the primary area where cracking begins is typically between the valves at the perimeter of the combustion chamber, which is on the extreme right side of the heads, and then propagates toward the spark plug hole and beyond, which may ultimately get near the "window area" at which point an entire corner of the head would be almost broken off. I've heard others knowledgeable in aluminum casting suggest that the primary cause of the head weakness/cracking is residual stress in the part due to shrinkage. A proposed solution for this large complex stressed casting was to provide huge risers over the combustion chamber to provide a very generous reservoir of molten material to allow back-filling of the mold as shrinkage occurs upon solidification/cooling, thereby reducing stress in the casting.

I've seen several heads that appear to have fissures or shallow stress cracks that never move or develop into anything more, but perhaps are reminders of past stress that occurred during casting?

sportsterpaul wrote:Patrick also explained how the patterns were made by the foundry, not at the factory. So the fact that the two heads on the left on the original post are a tiny bit different makes sense. If Harley changed foundries, I could see the first one not giving up their patterns, so the "E" foundry made a new pattern to the older design and ran those, maybe in 1953. Then in 1954 the change came in, and the "E" foundry made or modified the core and changed the pattern to make the small-window head with the different fin on top.

Not being nit picky but the heads as cast are absolutely identical with the exception of the casting number where one has "underscore 1" and one has "encircled E". The difference in the combustion chamber is from a machining operation occurring after the fact.
User avatar
wz507
 
Posts: 356
Joined: Fri Dec 11, 2015 9:13 pm

Re: Heads Up – K Model Head Quandary?

Postby LDB » Mon May 01, 2017 3:23 pm

wz507, I have a few questions for you if you don't mind. It looks like to me the head you measured was milled or cut slightly in the valve area on both sides with the typical milled circles for more clearance above the valves; not as deep or pronounced as some but one can see it in the picture with a definite radius cut in the casting.

Were all the heads made that way at the factory or was it done by some customer at a later date? Or in other words, were the heads, as they came new and installed from the factory, originally, the same depth across the flat area, e.g., A to B, or were they all at an angle across the flat area above the valves, whether as cast or cut? I fail to see why there would be an advantage to having the clearance at an angle, unless it is required to keep a tilted valve from striking the head. Otherwise, I see no advantage. And if an angle were required, why wouldn't the factory have the patterns made to cast these areas at an angle? Thanks for your reply.
LDB
 

Re: Heads Up – K Model Head Quandary?

Postby sportsterpaul » Mon May 01, 2017 4:00 pm

Oh, when I said tiny differences, I meant that in the bottom picture of the original post, it looks like the middle head has a little ridge at the top right where the left head does not seem to have that ridge-- maybe its just the light.

Yeah, casting stresses can cause cracks. I talked to a guy that made strain gauges and he told me a story where he did load testing on a casting and the casting broke. The engineer came over and told him he tested it wrong since the simulation on the computer showed the part would not break. The engineer copped an attitude that the testing was wrong. Turned out it was just what you said-- internal stress in the part from hardening in the mold added to the design load and the part would break. The simulation did not assume any pre-existing loads in the part.

I do think Patrick is onto something too, that thermal expansion can cause the cracks, especially for engines that got overheated. Its probably a combination of both problems. Patrick also pointed out how the foundry adds sprues and vents and other things. The tall sprue both puts pressure on the metal at the bottom and as you noted, gives some molten metal that can flow into the mold as it cools. Its a real art to do sand casting right. These kind of headaches with aluminum is one reason Harley may have decided to use cast iron heads for the Sportster.

I have to assume high-lift cams might cause some clearance problems, but if the parts book just shows one part from 54 to 56, maybe any machining differences were done by motor builders. I suspect you have will have to do just what you are doing-- making measurements and checking clearance with clay.
sportsterpaul
 
Posts: 240
Joined: Mon Jul 06, 2015 2:17 am
Location: Sun City Center, Florida

Re: Heads Up – K Model Head Quandary?

Postby wz507 » Mon May 01, 2017 4:05 pm

LDB wrote:wz507, I have a few questions for you if you don't mind. It looks like to me the head you measured was milled or cut slightly in the valve area on both sides with the typical milled circles for more clearance above the valves; not as deep or pronounced as some but one can see it in the picture with a definite radius cut in the casting.

Were all the heads made that way at the factory or was it done by some customer at a later date? Or in other words, were the heads, as they came new and installed from the factory, originally, the same depth across the flat area, e.g., A to B, or were they all at an angle across the flat area above the valves, whether as cast or cut? I fail to see why there would be an advantage to having the clearance at an angle, unless it is required to keep a tilted valve from striking the head. Otherwise, I see no advantage. And if an angle were required, why wouldn't the factory have the patterns made to cast these areas at an angle? Thanks for your reply.


LDB,

Yes, as you note one of the heads was virgin and unmachined (the left head was used as cast - I believe for a K or KH), one was spot machined in the valve areas (the middle head - I believe for a KK), and the right head was more fully machined (I believe for a KHK). All the heads were cast with the combustion chamber reflecting the tilting of the valves toward the combustion chamber (~ 3 deg?) as well as listing sideways (~ 5 deg?), so no, there was never a head having a combustion chamber cast or cut parallel to the gasket surface. The angularity of the combustion chamber was done to maintain a fixed clearance between the valve and cylinder head as is common in most engines. You can see in the Table I posted above that with respect to the IN valve the difference in depth between the high and low side (A to B) of the valve is on the order of 0.080 to 0.100”, so the chamber has to be tilted accordingly to maintain clearance.

The heads were cast to reflect the valve angle and the machining marks you reference were performed at the same complimentary angles simply to clean up the combustion chamber slightly for either larger valves or higher lift cams. But as my measurements suggest, unless my subject cylinder has severely sunken valves, even the early unmachined stock head (the left head) as cast could accommodate the KK/KHK cam lift. Hope this answers your questions. Thank you.

Hopefully those with superior knowledge will chime in and correct or clarify anything I may have muddied up.
User avatar
wz507
 
Posts: 356
Joined: Fri Dec 11, 2015 9:13 pm

Re: Heads Up – K Model Head Quandary?

Postby LDB » Mon May 01, 2017 5:39 pm

Thanks for your reply, wz507. So if all cylinder heads were not cast parallel to the gasket surface then that means the round quench area recess above the piston must be at an angle?

I can see how there might be some benefit if the recess above the valves were tilted toward the cylinder bore, but both valves tilted one way to the side makes no sense to me unless the valve is at such an angle that it would hit the head if a matching slant was not maintained in the head. That would have to be quite a slant of the valves.

Have you ever measured a clean head that has not been touched in the valve area to confirm they are at an angle or have all the ones you've measured but cut deeper with the milled circles to one depth or another? Thanks.
LDB
 

PreviousNext

Return to K, KK, KH, KHK

Who is online

Users browsing this forum: No registered users and 14 guests