Cylinder head - What can easily be achieved


The engine build I recently covered in previous Mini Magazine issues developed into a bit of an ongoing 'project'. During a discourse with the editorial staff at Mini Mag, it was decided the build feature presented an ideal opportunity to demonstrate just exactly what such an engine build is capable of in days where it's generally believed you have to have an all-singing, all-dancing 1380cc engine to have an enjoyable road burner - leaving those with very limited budgets a little depressed.


So the idea was to finish the engine off using a relatively 'mild' specification to maximise drivability. The camshaft used was the fantastically versatile Swiftune Racing SW5 profile that provides drive from nowhere up to 7,000rpm - depending on the finished spec - with minimal hydrocarbons for emissions testing. We're looking at the most used rpm band of 1,500-5,500rpm. Thrashing an engine to death to achieve any progress doesn't make for a pleasant driving experience on the road except for the masochistic.


To complement the camshaft, a suitable cylinder head was needed since - providing the engine has been decently assembled - an engine's power output is mainly developed from the camshaft and cylinder head. Volunteering my services to carve a head out for this engine instead of blagging one from another supplier, I took the opportunity to deal with a number of questions about cylinder heads I hadn't yet covered 'in public'. These centered around questions such as 'which head is best of the standard ones?' and 'I have an old standard S head - is it worth fitting on my MG Metro engine' and 'I've been told 3-angle seats cut into the head dramatically improve flow - is this true, and is it worth doing on a standard head?' and so on.


The plan then was to start out by flow-testing a bog-standard MG Metro head then working through a number of practical and often-used modifications and re-flow testing at each stage to gauge effects. The stages would be -


1 - Test with 30-deg back-cut on standard valves and reduced seat width on standard seats in the head

2 - Test with race-profile, waisted valve stem valves on standard seats for comparison

3 - Test with 3-angle seats cut into head and standard valves

4 - Test with as above but with standard valve with 30-deg back-cut and reduced seat width

5 - Test as above but using race-profile valves for comparison

6 - fully modify head and re-test


As outlined above, I selected the MG head as a starting point because it is the most efficient factory-fitted head used on the A-series and nobody - to my knowledge - has produced any useful comparative data. A consequence being folk still believe the old Cooper S head is the one to have. This simply isn't so. Leyland/Rover went to great lengths to ensure the MG Metro produced plenty of useable power since it's opposition at the time in the 'hot hatch pack' were running away with the performance crown. Partly through the seriously sporty (for Leyland) camshaft profile used - a cross between the 997 Cooper inlet profile and the old 731 fast road exhaust profile - and partly through very careful preparation and machining of the cylinder head; especially the inlet valve throats.


All previous A-series production cylinder heads have the valve throats simply bored out and the valve seat cut in with some small consideration made to the chamber side of the seat by way of a 'leveling' radiused edge cut around the seats' perimeter. The MG head, however, have the throats machined using a tool that cut the valve throat up to the underside of the valve seat to a special profile - a smooth concave curve terminating in a shallow convex ring around the inner seat edge. Thus forming a slight 'choke' effect and eradicating any sharp edges that inhibit good airflow. This choke effect is used in all manner of high performance engines; I have to say I was surprised to find this in an A-series engine the very first time I stripped one of these heads down. I won't go into a dissertation on how and why this choke effect works since space is limited; suffice to say it works in certain circumstances, and the A-series MG engine is one.


Of immediate interest is the fact that direct comparisons across the entire test reveal seemingly very small gains. This is something I alluded to earlier - the gains made by modifying cylinder heads aren't in the 'stupendous' region many folk seem to believe. An average gain of 10% is considered extremely good. I say 'average' here because the actual gains at each measurement increment vary and it's the nett result that's important, not a huge increase at one particular point. Having said that there are flow figures achieved at certain lift points that indicate a good head for certain applications.


A decently modified cylinder head can make a big difference to performance, but it isn't an easy thing to do. The main factor that makes this somewhat difficult is getting the air to perform at it's best at all speeds. Air doesn't much like flowing in any other direction than straight lines. To further complicate matters this single-minded approach by air to direction changes gets stronger the faster the air is moving.  When a head is modified, it needs to be done in such a way that the air is persuaded to deviate from its single-mindedness at all speeds. A head for full-race applications will have slightly different needs to a road engine and whilst it's reasonably easy for a skilled head modifier to achieve decent results for either one, the most skilled are those that can combine both to achieve perhaps the pinnacle of results. Improving the flow in as many areas as possible is critical in developing more power since higher engine rpm means higher air speed, consequently the more difficult it is to persuade the air to flow from the port, round the right-angle bend onto the back of the valve and eventually out into the chamber. Inexperienced or clumsy modifications - particularly in the port throat and that 90-degree bend - can loose airflow instead of gaining it. It is also interesting to note that no matter what the modifications the improvements don't really start showing until the valve is open more than 0.100"/2.54mm. Also of note is the fact that both exhaust and inlet valves account for a similar obstruction value - around 4-5cfm.


When scrutinising the results very carefully and comparing each improvement against the base data of the standard MG head, it's interesting to note that it's the exhaust port flow that sees the biggest overall gains. The gain on the inlets alone doesn't explain the sort of power increases achieved with a modified head. It would suggest the more efficient, and therefore less restrictive, exhaust ports have more of an effect through allowing better breathing via more complete blow-down of the cylinder. In other words, a greater percentage of the burnt, power producing fuel/air charge is evacuated causing the cylinder fill with a greater volume of fresh in-coming fuel/air charge - improving the engine's volumetric efficiency (it's capability to fill the whole cylinder with fresh fuel/air charge). The greater efficiency also reduces power loss caused by the piston having to physically push the exhaust gas out.



Valve lift   Test 1   Test 2   Test 3   Test 4   Test 5   Test 6   Test 7   Orig S

    (in.)                                                                                              head


0.050        23.8      24.7     24.0      24.5     24.8      24.7      24.4      19.5              

0.100        42.1      43.0     42.2      43.3     43.0      42.7      45.7      43.0

0.150        58.3      63.2     63.2       59.1    63.6       63.4     67.3      53.1

0.200        75.3      80.6     81.6       75.2    81.6       81.6     87.2      69.1

0.250        89.8      94.7     96.2       87.9    95.3       95.6    102.1      85.5

0.300        99.5    103.5   103.2       97.8   104.8     104.6   113.8      93.4

0.350      105.4    107.8   108.1     105.2   108.0     108.2   119.4    102.1

0.400      108.7    109.5   109.1     108.5   109.4     109.2   123.5    106.8

0.450      109.8    109.1   109.3     108.8   108.9     108.9   127.4    109.0

0.500      110.1    110.3   110.2     109.4   110.0     109.9   128.8    109.0

0.550      111.4    110.2   110.8     110.0   111.1     111.0   129.0    108.0



0.050        19.1      20.6      21.2      19.0      21.0      20.0      18.8     14.3

0.100        33.2      37.2      37.0      33.4      39.1      38.2      33.0     32.8

0.150        47.8      51.3      50.3      43.9      51.7      50.0      62.9     40.4

0.200        56.3      59.3      58.1      55.4      59.6      57.7      63.9     49.4

0.250        63.6      66.0      64.3      63.2      66.2      64.8      71.9     58.1

0.300        68.9      70.7      69.2      68.6      71.2      69.9      78.0     62.4

0.350        72.7      74.1      73.0      72.5      74.2      73.8      83.2     66.0

0.400        75.0      75.8      75.4      74.2      75.8      75.8      87.5     68.0

0.450        76.1      76.6      76.8      75.0      76.3      76.9      90.6     68.0

0.500        76.8      77.2      77.5      75.5      76.8      77.0      93.1     67.0

0.550        77.0      77.6      77.9      75.7      76.9      77.3      94.8     67.0


NOTE: All tests carried out at 25-inches pressure drop. The figures given are the averages taken from all the ports on one head. For comparative purposes the standard ports were flow tested with no valves - results were inlet 115.0cfm and exhaust 79.0cfm. Although the S head was tested at different premises some years back, the same machine was used so is reasonably relevant.


Test 1. This represents the 'as standard' head. I tested seven such heads and selected the most 'average' one. There are differences from head to head, some flow quite substantially less with the odd few flowing a little more - but not by much. This variation is brought about by the very inconsistent castings with core shifts going in all directions. Experienced head modifiers know what to look for to select the best castings to get the best 'as cast' performance from a standard head and gain the biggest improvements when modifying them. Unfortunately the A+ castings are nowhere near as good as the older pre A+ items; these good 'old' castings are becoming more and more difficult to get hold of. The valve seat widths are quite huge - some 0.090-in. on the valves and 0.080-in. on the head, more for longevity than anything else. 


Comparing these figures with those of the original S head figures in the last column graphically illustrates the improvements made to the MG head by careful attention to the valve throat detailing and as cast port shapes mentioned previously - particularly in the sub-0.350-inch lift area where most standard-type road cams operate, the valve has a significant effect on flow because it's near the valve seat and bearing in mind that a 10% increase is a really good step forward which the inlet pretty-much averages. The exhaust, however, averages very nearly 15%.


Test 2. Here we have the standard head tested with a 30-degree back-cut on the valve to improve flow across the back of the valve head and to reduce valve seat width to approximately 0.060". This simple modification that takes barely minutes to execute improves flow by an astounding 6-ish% on the inlets and exhausts - as much as the MG gained over the S head.


Test 3. A race-specification valve developed to give maximum flow/minimal obstruction complete with waisted (narrowed) valve stem, near flat back and 30-degree back-cut before the seat was fitted to see if it would make any further gains in this application. The answer is a resounding no - the results are near identical to that of the modified standard MG Metro spec valve.


Test 4. So to the famed and much acclaimed 3-angle seats cut into the head. This was the typical 25-degree top cut, 45-degree seat angle at 0.060-in. wide, and 65-degree under cut widely used. This test employed the standard, unmodified MG valve complete with 0.090-in. wide seat. The result will astound many since the overall effect in this instance is marginally worse than the entirely standard set-up!


Test 5. To try and re-dress the loss and hopefully regain some faith in the long-held belief that 3-angle seats are the crowning glory the modified MG valve was fitted and tested. Hallelujah! Faith restored - except note how very small the gains are in comparison to test 2.


Test 6. For consistency's sake the race valve was again tested with astonishingly similar effects, i.e. it made nearly no difference whatsoever.


Test 7. Finally to the head in modified state I prepared (and fairly typical of an average Mintec modified road head) for road use with a reasonably mild cam (the SW5), so we're looking for decent gains up to 0.450-in. As can be seen from the test data - the standard un-modified head does OK up to the 0.350-in. level, then hits a bit of a brick wall even with the 3-angle and 30-degree back-cut treatment. This is where a decently modified cylinder head makes the difference by improving flow beyond this point whilst not loosing - even hopefully gaining - at the lower lift regions. Muted earlier, the A+ heads are not the best for maximum gains, and in particular the MG heads as they have this carefully profiled intake throat immediately before the valve seat and different core centres when casting. Even so, the test results show that the goals have been achieved with respectable results. More importantly, no losses!



The MG Metro head is really pretty good in standard trim - certainly a major step forward compared to the original 'S' head. Three angle valve seats adversely affect flow unless used in conjunction with re-profiled valves having the 30-degree back cut on them. The modified standard valves with the 30-degree back-cut are amazingly effective at increasing airflow for little expenditure. However, a reasonably well modified head pays dividends all over, especially when up-rated cams are used.



Useful part numbers:

Note - All modified cylinder heads now come converted for lead free as standard. For full spec on heads, see inventory. Specific application heads can be done to special order.


MSE3           Min Tec pre-92 large-bore road-rocket spec head,

                    race quality components, 35.6mm/29.5mm dia. valves.

MSE4           Min Tec post-92, including SPi, large-bore road-rocket head,

                    race quality components, 35.6mm/29.5mm dia. valves.

MSE7           Min Tec TPi road-rocket spec head, race quality

                    Components, 35.6mm/29.5mm dia. valves.

C-AHT135    Pre-'92 large bore road-rocket spec head, A+ standard

                    quality components, 35.6mm/29.5mm dia. valves.

C-AHT136    As C-AHT135 except for Twin point injection engine applic.

C-AHT88      Min Tec Stage 3 small-bore head, race quality

                    components, 31mm/26.4mm dia. valves.

C-AHT55     1.464"/37.2mm S/Steel inlet valve, large-bore.                                        

C-AEG544   1.401"/35.6mm S/Steel inlet valve, large-bore.

C-AEG543   as C-AEG544, but with triple-groove collet fitting.                             

C-AEG569   1.311"/33.3mm S/Steel inlet valve, large-bore.                                       

C-AEG588   1.218"/30.9mm S/Steel inlet valve, small-bore.                                     

C-AEG107   1.220"/31.0mm S/Steel exhaust valve, large-bore.

C-AEG106   1.161"/29.5mm S/Steel exhaust valve, large-bore.

C-AEG105   as C-AEG106, but with triple-groove collet fitting.

C-AEG587   1.040"/26.4mm S/Steel exhaust valve, small-bore.

C-AEA526    Nominal 180 psi dual valve springs, road/rally use up to

                    0.495" lift/7,500 rpm continual use.

C-AEA527    Nominal 240 psi dual valve springs, rally/race use up to

                    0.570" lift.

C-AEA528    Alloy top cap set, hard anodised, very light, single groove

                    collet only.

AEA653        Steel 'W' type top caps, single groove collet only, need 8.

C-AEA654    Competition inner spring locator (no step).

AEA403       Standard lower inner spring locator, need 8.

SS4             Set of 4 spring spacer-shims, 0.080" thick.

88G459       Single groove collet, need 8 (come in pairs).

ADU4905    Valve stem seal, late 'sprung' type.

AEG327       Valve stem seal 'S' type without spring.