[ROVERNET - UK] 3500s head compatibility

Robert Thornton R.Thornton at adelaidecitycouncil.com
Mon Feb 20 23:04:36 GMT 2006


I appreciate what you say about this. However it seems to me there is
little point in choosing a 3.5 block over the bigger bore items. If what
Hammill says is right, then all two bolt mains blocks (i.e. virtually
all 3.5 blocks) will suffer fretting due to inadequate cap register
surfacing rendering these blocks scrap. They cannot be repaired. While
ARP studs help they do not absolutely overcome the problem. Rover fixed
it when it introduced the 94 mm X bolted blocks in 1995, i.e 4.0 and

The problem with these larger blocks is the one of cracking. BUT there
is a solution in the form of top hat / flanged liners which effectively
overcomes this problem. Unlike the two bolt 3.5 / 3.9 blocks the larger
ones are salvageable. The 4.0 / 4.6 blocks seem to be the lesser of two
evils because they can at least be fixed, whereas the 3.5s are history
once the bearing caps start to fret.

A used 4.0 / 4.6 would be ok if you were going to re-sleeve it using
flanged liners. Then even if it had cracked the new liners would prevent
coolant escaping into the combustion chamber or crankcase. Hammill
claims this is a 100% effective cure for a cracked block, albeit a
rather expensive one.

I suppose the worst case scenario is a 3.9 which has two bolt mains and
a 94 mm bore - 50% chance of the main caps fretting, 25% chance of the
bores cracking (according to DH's calculations). 

Having said that my own 3.9 Disco, now ten years old and used as my
daily transport, has run faultlessly, never uses a drop of coolant and
has never overheated, despite the 40 degree plus C temps we often
experience here in summer. I don't know what condition the mains are in
because I've never had it apart. But I do keep the cooling system in
good order and routinely take the radiator out and have the cores
cleaned thoroughly.


-----Original Message-----
From: rovernet-bounces at lyris.ccdata.com
[mailto:rovernet-bounces at lyris.ccdata.com] On Behalf Of Kent Kinard
Sent: Monday, 20 February 2006 3:07 PM
To: rovernet at lyris.ccdata.com
Subject: Re: [ROVERNET - UK] 3500s head compatibility

Hi Rob,
OK, a new 3.9 at the right price, maybe.  No chance of it having been
overheated and thus developing main cap or sleeve problems.  But a used
3.9, never.  You can't be sure it wasn't overheated and there just isn't
enough aluminum in the cylinder wall anyway(.120"/3.0mm with perfect
cores). If you overbore a 3.5 and use thinner sleeves (.055" vs. .091)
you can have .152-.156" wall thinkness with a 3.702" bore.

This being said,Hammill seems to think the top hat sleeve cures all
ills.  It's worth doing to a 4.6 or a 4.0 with decent cores (although a
4.0 with good sonic test results would be a rarity according to
Hammill).  Unless you can find the markings on a 4.0 block or determine
core shift, it still might not be worth the machine work costs.  Too
many question marks.  I'll just say I won't buy used 3.9's or 4.0's at

Since I already have a very low mileage 4.0 with a bad sleeve (and
presumeably a cracked cylinder behind that sleeve, I'll go ahead and
have the sleeves removed and the block sonic tested and try the top hat
sleeves to see if they really will hold up.  I wish I had read the book
before I had my 4.6 done.  I may have wasted my money having it
resleeved without using the top hat liners.  It had a loose sleeve also.

Those with 3.702 bore engines that have never overheated will have
relatively few problems, but most of us on Rovernet are looking to
re-engine an older vehicle or rebuild one we already have.  If you know
the service history of a vehicle and the vehicle was wrecked, the engine
may be one that has never and will never give problems, but statistics
on the 3.9 are not in your favor.

Given the use we will put them to and the questions surrounding factory
large bore blocks, we are better off using 3.5 blocks, particularly if
they have no history of overheating or mysteriously losing coolant. Even
an unkown 3.5 block with no signs of water jacket erosion, would be a
better bet statisically than an unknown 3.9 block. The street
environment is definitely harder on engines in some ways than a race
environment.  I have had two 3.5's that were chronically overheated and
developed cracked cylinders behind the liners.

We haven't even talked about the problem of ovality at the bottom of the

Maybe the IOE engine wasn't such a bad deal.

Kent K.

Robert Thornton wrote:

> Kent wrote : 
>>>you are MUCH better off starting with a 3.5 block, either early or
> late, than you would be investing in a factory 3.9, used or new. I 
> will never again buy a 3.9 for any reason. I would only buy a 4.0 if I

> knew the vehicle and it's service history from new and intended to fit

> "top hat" sleeves. <<
> Hi Kent
> What's the difference between investing in a new / used 3.9 or 4.0 
> block? They use the same bore, same liners. OK, the 4.0s are cross 
> bolted whereas the 3.9s are not, although they have provision for it 
> and the later interim 3.9 blocks have the longer crank to drive the 
> oil pump. But this does not influence the porosity issue.
> Des Hammill seems to have scared a lot of people off 94 mm blocks.
> There's been much debate about this on the Land Rover Owners forum
> http://www.lro.com/nav?page=lroi.message.list&section=MESSAGE_READERS_
> FO
> Consensus of opinion, if I interpret it rightly, is that yes, a 
> significant number of 94 mm block do suffer this problem, particularly

> if they are allowed to overheat. But a lot do not ever become porous, 
> and these may be in the majority. Some other manufacturer's alloy 
> engines can also have problems with heads / blocks cracking if they 
> are allowed to seriously overheat.
> Moral of the story seems to be keep your cooling system in top notch 
> order, regularly remove and clean out the cores of radiators and 
> closely monitor the operation of radiator hoses, thermostat, water 
> pump and cooling fan. In the case of P38 Range Rovers a complete 
> re-think of their cooling system is really warranted for all the 
> reasons Hammill identifies.
> Rob
> _______________________________________________
> rovernet mailing list
> rovernet at lyris.ccdata.com
> To unsubscribe, go to this web page, look near the bottom and follow
> http://mailman.nipltd.com/mailman/listinfo/rovernet
> Back-up list and photos at:
> http://autos.groups.yahoo.com/group/Rover_net/

rovernet mailing list
rovernet at lyris.ccdata.com
To unsubscribe, go to this web page, look near the bottom and follow
Back-up list and photos at:

More information about the rovernet mailing list