Last & Best of the Piston Engine Fighter Aircraft.

[Tommy Cookers]

... Napier chose W12 to avoid casting difficulties of a V12 (later they rejected the job of 'emulating' the Curtiss V12)

T-C, as for Napier's finding any"difficulties" with casting, somehow their next aero-engine was an X-crankshaft 16 cylinder!
(Nor did Frank Halford's later 16 & 24 cyl' designs, appear to give Napier major "casting difficulties").

I wrote 'difficulties of casting a V12' not as you imply 'difficulties with casting'
some V12s at the time of the Lion even having eg crankcase and 12 water jackets etc as one big casting - difficult

Rowlege (afaik it was made known at the time) chose more conservative (my word) construction
the Lion had 12 cylinders separate from each other and from their water jackets and those separate from the heads
the 16 cylinder Cub had separate cylinders with fabricated water jackets
the Halford engines had cylinders separate from each other and separate from the 2 separate crankcases

Napier refused the task of 'building a new engine with (Curtiss) D-12 qualities' so causing the RR Kestrel etc ....
but then built a separate-cylinder V12 that was unsuccessful

[J.A.W.]

... Napier chose W12 to avoid casting difficulties of a V12 (later they rejected the job of 'emulating' the Curtiss V12)

T-C, as for Napier's finding any"difficulties" with casting, somehow their next aero-engine was an X-crankshaft 16 cylinder!
(Nor did Frank Halford's later 16 & 24 cyl' designs, appear to give Napier major "casting difficulties").

I wrote 'difficulties of casting a V12' not as you imply 'difficulties with casting'
some V12s at the time of the Lion even having eg crankcase and 12 water jackets etc as one big casting - difficult

Rowlege (afaik it was made known at the time) chose more conservative (my word) construction
the Lion had 12 cylinders separate from each other and from their water jackets and those separate from the heads
the 16 cylinder Cub had separate cylinders with fabricated water jackets
the Halford engines had cylinders separate from each other and separate from the 2 separate crankcases

Napier refused the task of 'building a new engine with (Curtiss) D-12 qualities' so causing the RR Kestrel etc ....
but then built a separate-cylinder V12 that was unsuccessful

Rowledge was using typical practice for the Lion's individual cylinders, though the en bloc cyl' heads
with fully enclosed valve-gear, (liberally supplied with lubricant as coolant), were innovative.

Of course Halford's & Rowledge's later air-cooled inline engines (as per radials) - utilised individual
cylinders, (just as inline 4 racing motorcycles such as Remor's Rondine/Gilera/MV Agusta also did) but
Napier's fairly sophisticated castings for the Sabre's liquid cooled cylinder blocks - certainly did not...

[Tommy Cookers]

.... Napier's fairly sophisticated castings for the Sabre's liquid cooled cylinder blocks ...

it has been said that ....
'Tony' Rudd (a 'Sabre-lover'), having designed the over-heavy and over-bulky H16 BRM engine ....
then complained that the big castings were much thicker and so much heavier than designed
the foundry made the patterns (thickness) to give a successful cast ie the design as BRM-drawn was uncastable
all this after the first design concept (2 camboxes and 6 camshafts) had been abandoned by BRM as uncastable
abandoned for the bulkier/heavier design with 4 camboxes and 8 camshafts and increased crankshaft spacing
(minimal crankshaft spacing being vital to CG height as the lower bank height was dictated by exhaust pipe run)

btw
by 1955 the V16 BRM was reliable, very high-revving, and very powerful
(the 1954 2.5 litre F1 was decided before the 1952 WDC replacement of 'V16 friendly' F1 by F2 etc)
in late 1958 the FIA demanded 1.5 litre F1 from 1961
how didn't BRM even try an NA'd V16 ? (surely a winner from day 1)
... the hot ticket in 1965 was the transverse V12/central power takeoff Honda- BRM already had this in the V16

and btw ....
the 1966 '3 litre' F1 that (loosely) lasted 4 decades was announced in 1964 - and generous to the supercharged option
if BRM had used an NA version of the V16 1.5 litre from 1961 they could have also .....
won from day 1 in 1966 by reinstalling supercharging - eg initially a better-regulated 4 bar single-stage
(we might see turbocharging in F1 as a better-regulated 4 bar single-stage)
BRM had intended boost regulation by RR VG vanes (never implemented but interstage throttling later used instead ??)
in the 1968 SA GP demo they ran the V16 (with early type larger induction system) to 13000 rpm and c.780 hp
RR had said the V16 valve timing (less than 300 deg) and lift was a limiting factor (though safe to 190000 rpm)

NOTES TO SELF
BRM didn't stretch its 2.5 litre four for Intercontinental formula and 1961 F1 weight limit discouraged a related twin
Coventry Climax DID design an FPF ((94 x 90 2.5 litre) destroked to 94 x 54 for 1.5 litre !!

the point of the H16 was to use the 1962-5 winning 1.5 litre V8 innards
even without the other H16 design faults this was rather backing the wrong horse
enlargement (mostly by stroking) of 1.5s (originally for Ginther's US connections) showed the 1.5s rod ratio was wrong
ok the 64 valve H16 was maybe less ratio critical (also BRM and Weslake (related) 48 valve V12s) than 2 valve versions
but even without its faulty exhaust port divider design the 64v H16 wouldn't have beaten the DFV ?
(though at that time weight limits helped H16s by allowing them magnesium monocoques but not so DFV cars)

the H16 BRM /DFV Lotus 49 comparison I likened to a Sabre vs V12-power fighter plane comparison - 5 years ago !

[J.A.W.]

.... Napier's fairly sophisticated castings for the Sabre's liquid cooled cylinder blocks ...

it has been said that ....
'Tony' Rudd (a 'Sabre-lover'), having designed the over-heavy and over-bulky H16 BRM engine ....
then complained that the big castings were much thicker and so much heavier than designed
the foundry made the patterns (thickness) to give a successful cast ie the design as BRM-drawn was uncastable
all this after the first design concept (2 camboxes and 6 camshafts) had been abandoned by BRM as uncastable
abandoned for the bulkier/heavier design with 4 camboxes and 8 camshafts and increased crankshaft spacing...

... the hot ticket in 1965 was the transverse V12/central power takeoff Honda...

and btw ....


...the H16 BRM /DFV Lotus 49 comparison I likened to a Sabre vs V12-power fighter plane comparison - 5 years ago !

T-C, your allusion reiterated after 1/2 a decade - remains more illusory - than you may imagine.

That lumpen H-block BRM bears only a very superficial resemblance to F. Halford et al's elegant work.

Cosworth's engines really started to make power once they'd combined Honda's DOHC 4V head with
liquid cooling, plus (Sabre-like) robust full crank-support crankcase & quill-shaft expediencies..

Also, K. Duckworth & M. Hewland were sufficiently impressed by sleeve valve advantages to build an
experimental 500cc single cylinder test unit (with potential as a prospective DFV successor) which
apparently ran reliably, & easily handling thermal stresses while running an S.P.O. of 144hp/litre...

R/R could not convince de Havilland or Hawker designers to accept a 36 litre V12 Griffon as a viable
replacement of the (also 36 litre) H24 Sabre for their larger high-speed airframes effectively,
(neither on paper, or via actual in-flight trials).

So T-C - your imagined ("likened") comparison - really does not stand analysis.

[wuzak]

Cosworth's engines really started to make power once they'd combined Honda's DOHC 4V head with
liquid cooling, plus (Sabre-like) robust full crank-support crankcase & quill-shaft expediencies..

Huh?

What did Honda have to do with Cosworth?

And what did they use quill shafts for?

Also, K. Duckworth & M. Hewland were sufficiently impressed by sleeve valve advantages to build an
experimental 500cc single cylinder test unit (with potential as a prospective DFV successor) which
apparently ran reliably, & easily handling thermal stresses while running an S.P.O. of 144hp/litre...

Then why didn't they pursue it? There were no regulations requiring poppet valves back then.

R/R could not convince de Havilland or Hawker designers to accept a 36 litre V12 Griffon as a viable
replacement of the (also 36 litre) H24 Sabre for their larger high-speed airframes effectively,
(neither on paper, or via actual in-flight trials).

de Havilland had two "super Mosquitoes" on the drawing board. One was to use Sabres, but they would not be able to get a sufficient supply of those engines, so it was dropped.

The second was to be powered by the Griffon, but the performance increase over the in-service Mosquito was not enough to warrant further development.

de Havilland did not try to replace a Sabre with a Griffon. And it was likely the Air Ministry shelved these projects, rather than de Havilland.

Hawkers were told to use the Griffon as one of the engine options for the Tempest and/or Fury. Since the others were the Sabre and Centaurus, any advantage the Griffon would have provided in aerodynamics was lost.

The Griffon was lower powered than the other two. Two Griffon models were to be trialled - a single stage engine and a two stage engine. The latter may have provided better altitude performance, but was majorly disadvantaged and lower altitudes, where the Sabre and Centaurus shone.

[wuzak]

btw
by 1955 the V16 BRM was reliable, very high-revving, and very powerful
(the 1954 2.5 litre F1 was decided before the 1952 WDC replacement of 'V16 friendly' F1 by F2 etc)
in late 1958 the FIA demanded 1.5 litre F1 from 1961
how didn't BRM even try an NA'd V16 ? (surely a winner from day 1)
... the hot ticket in 1965 was the transverse V12/central power takeoff Honda- BRM already had this in the V16

The V-16 was long and heavy. It did not fit well with those small mid-engined F1 cars.

I can't see how Honda's transverse V12 was "the hot ticket", since it won only once. There were 10 races in 1965, Honda competed in 8 and had 14 starts, retired 8 times, finished 6th twice, 5th once and 1st once, and had two finishes outside the top 6.

There is some benefit to the transverse layout - such as greater drive line efficiency (no need to turn the power 90°). But there were some packaging difficulties.

Notably, Honda's V12 for the 3L formula was not transverse.

Ferrari won both championships with a V6 in 1961, BRM both in 1962 with a V8, Lotus both in 1963 with a V8, Ferrari won both with a V8 and Lotus won both with a V8 in 1965.

Note that Ferrari ran a V12 at selected rounds in 1964 and 1965, with limited success.

[J.A.W.]

Yeah wuzak, Cosworth carefully examined the extant DOHC 4V racing mills winning in the early`60s,
& knew that Honda had the 'pent-roof 4V' design working best, much better at high rpm than their 2V,
with heavy valves & stiff spring requirements that were proving problematic, metallurgy-wise.

Air-cooling was an 'old man Honda' fixation, but that hegemony meant nothing to Cosworth.
Additionally, they'd taken note of the full-support crankcase also being needful, due to shortcomings
shown in all palliatives used with production-based upper-crankcase crankshaft fixings.

Unlike BRM, where the subtle sophistications engineered into H-16/24 mills by F. Halford et al, were
missed/ignored, Cosworth also applied the 'torsion bar' tech seen in the valve train of the Sabre,
in their DFV ( & others) to likewise mitigate shock-loadings..

Why the Duckworth/Hewland sleeve-valve was not seen, I'd postulate it was cruelled by the DFV's
success as a production customer unit, capable of ongoing incremental power development that
kept it competitive, in the traditional British way of business 1st.

As for the failure by R/R to substitute their 'old school' fall-back V12 Griffon for (either their own
self-cancelled Vulture, or) the Sabre, for Hawker/de Havilland use in a large-fast twin, that was due
to its fundamental inability to develop sufficient power for high-speed cruise, & in the big Hawker
fighters, neither max-power, nor high power-setting endurance, to compete with Sabre in wartime
/post war hard use, or the big Brit Centaurus in peacetime/combat-exigencies at low-level...

[Tommy Cookers]

.... Cosworth's engines really started to make power once they'd combined Honda's DOHC 4V head with
liquid cooling ....

not so .....

eg the Cosworth SCA was an outstanding engine with a VIA of zero (and a freakishly high bore:stroke ratio)
eg in Honda's crushingly dominant years their GP motorcycle engine's VIAs were 76 degrees
(conspicuously the Cosworth DFVs were 32 degrees of course - and the Weslakes were 30 degrees)

only in later years when Honda increased the bore:stroke ratios was their VIA reduced at all
eg the 1966 500cc 58 x 47 4 cylinder motorcycle and its parent 1964/5 1500cc V12 F1 car used 56 deg

big VIAs were universal in racing pre-Cosworth&Weslake
additionally Honda's 44mm bore was presumably ok for combustion speed - but a DFV bore size wouldn't be
(and as per their SAE paper Honda used low octane fuel, presumably so to combust readily)
eg the 1.5 litre Coventry Climax V8 went to 73mm bore but kept the 62 deg VIA (even with 4 valves)
similarly everyone else who couldn't make 4 valves work well eg 60s BRM and 50s Norton/Weslake etc research
(and MB designing for 1954 F1 couldn't make their choice 4 valve work - so went desmodromic to enable a 2 valve job)

Cosworth and Weslake small VIAs apparently sacrificed valve size for better combustion
the world then copied this but in F1 used ever-higher bore:stroke ratios to give relief from the valve size issue
(both Cosworth and Weslake had bore size constraints when designing these small VIA F1 engines c.1966)

[J.A.W.]

Yes so, T-C.

Honda, being bent on air-cooling was forced to compromise on valve-angle, due to the need to
allow sufficient space for fins over the top of the combustion chamber, & around the spark-plug,
between DOHC cam-boxes - while Cosworth were not so constrained, & designed accordingly -
unlike Honda, whether on small capacity G.P. bikes - or for G.P. (F1) cars..

Funnily enough, although Suzuki used liquid-cooling on its racing 2Ts to beat Honda 4Ts in G.P. events,
& utilized it for their GT 750 road bike (& RE 5 Wankel) in the early `70s, they attempted to evade
the need for liquid-cooling of their hi-po DOHC 4V 4Ts, by dint of Wright-style close-finning, & excess
oil circulation, right up to the `90s, when they had to admit it was a dead-end, & long after Honda
had adopted closed-up valve-angles & liquid-cooling as needful, so ironically...

Even Yamaha, which had also adopted liquid-cooling for its 2T G.P. machines (& a DOHC Toyota),
resisted using it for their 1st 4V road bike, & paid the the penalty of over-heating/power-loss/
cracked heads/expensive warranty claims/re-design costs on their early `70s TX 500.

Honda hedged its bets by introducing 4V to road machines via SOHC singles - XL 250/350 'street
scramblers' (small-capacity 'soft-roader'/adventure bikes' of the early `70s) but when developed up
for MX performance with Yoshimura Racing parts, the Honda rocker arms were 'on a hiding to nothing',
being stuck between the 'rock & a hard place' - of a fat cam & stiff valve springs - something I know,
from experiential/empirical 'trials'... (& Honda then produced 2T MX bikes to compete).

[Tommy Cookers]

so you're now saying that Honda would have used small VIA water-cooled but for the needs of air cooling .....

so Cosworth were copying Honda when making something that Honda didn't make and hadn't ever made !!

[J.A.W.]

so you're now saying that Honda would have used small VIA water-cooled but for the needs of air cooling .....

so Cosworth were copying Honda when making something that Honda didn't make and hadn't ever made !!

My comment was that Cosworth closely examined the Honda DOHC 4V arrangement, & duly found
an advantage could be gained, by utilizing the ability of liquid-cooling to allow a closer valve angle.

Cosworth's experience with (for that time, ultra short-stroke) oversquare B x S engines was an artefact
of race-tuning the production Ford OHV - which was designed as such - for non-racing purposes.

Here below, 'from the horse's mouth':

https://www.grandprixengines.co.uk/Note_78.pdf

& as an on-topic addition, I'll also note that K. Duckworth was well aware of in-cylinder flow dynamics
having ah, - 'cut his teeth' - via big aero-engines (of sleeve valve type) at the Bristol works...

[wuzak]

Here below, 'from the horse's mouth'

You'll have to quote the actual section to which you refer.

[J.A.W.]

Here below, 'from the horse's mouth'

You'll have to quote the actual section to which you refer.

You can surely click 'n' read the three pages as linked, wuzak?

See page 2, paragraph 3 - describing Brabham's success in F2 with Honda in 1966, winning over
the previously dominant Cosworth SCA ,(which ironically enough, was mirrored via Brabham's F1
World Championship wins in 1966-67 - with ex-GM stock-block V8s cobbled-up by Irving-Repco
as OHC race-units per Cosworth's work with Ford - reliably enough to defeat the DFV!), plus BRM.

[wuzak]

Here below, 'from the horse's mouth'

You'll have to quote the actual section to which you refer.

You can surely click 'n' read the three pages as linked, wuzak?

See page 2, paragraph 3 - describing Brabham's success in F2 with Honda in 1966, winning over
the previously dominant Cosworth SCA ,(which ironically enough, was mirrored via Brabham's F1
World Championship wins in 1966-67 - with ex-GM stock-block V8s cobbled-up by Irving-Repco
as OHC race-units per Cosworth's work with Ford - reliably enough to defeat the DFV!), plus BRM.

It says that Duckworth concentrated on designing the FVA rather than improving the SCA in 1965-66.

It doesn't say anything about copying Honda or being inspired by Honda.

4V designs had been around for over 50 years by this point.

[J.A.W.]

I think perhaps you missed the point made in that treatise, wuzak, yes 4V had been
around that long, but other than in very large cylinder applications, where
individual valves & springs had proven to have a hard-short life in various
applications (esp' where valve gear was less than optimal, such as in the non-roller
cam-follower/rocker arms of the SOHC R/R V12 Merlin/Meteor), yet as such short life
was acceptable, in racing usage, so 2Vs made winning power, & 4Vs had shown no
performance advantage, (US Miller/Offenhauser blown engines excepted) so the real
performance of the Honda 4V, when translated from bike to car mills - was a big deal.

Honda resorted to palliatives such as torsion bar valve springs, but Cosworth knew
that using the same high rpm non valve-bounce powerband, really efficient cylinder
volumetrics, along with liquid cooling of the top-end, (& once they had the reliability
sorted), their DFV was def' going to be an F1 world beater..