Building a FIA appendix K Alfa Romeo 1300 Giulietta Sprint Veloce engine - Page 3 - Alfa Romeo Bulletin Board & Forums
 11Likes
Reply
 
LinkBack Thread Tools Display Modes
post #31 of 56 (permalink) Old 07-22-2018, 01:14 PM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Calibration of the flowbench

Just by using this simple flow bench it is possible to determine improved or worsened flow from porting work. But to get meaningful flow numbers the flow bench needs to be calibrated.

To be able to calibrate the flow bench, I machined a piece of plexiglas to create a “Helgesen plate” according to the dimensions outlined in David Vizard’s book, “Port and flow test cylinder heads” (In the Helgesen plate one blocks the holes one don't need for the moment). Since the alfa 1300 engine only has 74 mm bore, I had to place all holes closer to each other than presented in the book. This gave some wall flow and hole interaction effects, meaning that the flow calibration would differ depending on which way the plate was installed. I therefore also made a set of calibration plates with single, centered holes. The calibration can be performed anytime I lift of the cylinder head, to see that the calibration curves I created, still are valid or if there are problems with the equipment. For instance, if one of the vacuum cleaner pulls less air than usual (filter obstruction or net voltage variation) the effect is less pressure difference recorded on the manometer giving a false impression that the cylinder is flowing better than before…

By performing careful line fitting of the calibration curve, a matching line equation was created. This can be used to directly calculate flow from measured pressure difference instead of locating it on a calibration curve plot.

Never the less, I could not be sure of the quality of the calibration itself, so eventually I handed in one of the 1300 heads to a professional engine builder (Logmech) with a Superflow SF450 flow bench. The difference to my setup was relatively small, with my flow bench underestimating flow at most 6% at the highest 12 mm valve lift, and overestimating flow below 4 mm valvelift. Between 4-7 mm lift, flow was exactly the same. Obviously I used the results from Logmech to calibrate my calibration plates.

Conveniently, Logmech flowed a standard 2-liter Nord head for me many years ago. Logmech’s results match almost perfectly the flow figures published Jim Kartamalakis book. With the calibration process I have described, I think I can conclude that the flow figures I achieve on my simple flow bench are reasonably accurate.
Attached Images
   
horsewidower likes this.
DonMartin is offline  
Sponsored Links
Advertisement
 
post #32 of 56 (permalink) Old 07-22-2018, 04:32 PM
Senior Member
Gold Subscriber
 
horsewidower's Avatar
 
Join Date: Jun 2006
Location: Wilton Ca
Posts: 2,212
Very nice. Are you able to measure velocity.

We dyno'd engines with compression ratios ranging from 10.5:1 all the way up to 14:1. We found that over 12.5:1 there wasn't any advantage...and chalked it up to the piston crown interference with the mixture burn efficiency.

Again, Chrysler hemi experience, your experience may differ.


Sent from my iPhone using Tapatalk

Everyday is a school day. Learning something new all the time

View my '74 GTV restoration at:
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


View my '91 164s adventure at:
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


But wait there's more: View my 164LS thread:
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


You're kidding, another one?! 1984 GTV6 Maratona
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


You didn't...oh yes I did. '73 Berlina graduate course in rust repair. No thread yet, don't want to piss off the GTV

You didn't...oh yes I did II. Deposit down on another vehicle...stop the madness!!!
horsewidower is offline  
post #33 of 56 (permalink) Old 07-23-2018, 07:49 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Port dimensions

Before flowing anything, it is good to know what we start from.

To get a clearer view of the differences between intake ports of the three cylinder heads, I made casts in RTV. By adding several thin layers of RTV silicon with a brush, castings of the ports are made (make sure to wax the surfaces to simplify removal, before applying the RTV). For the 105 head I made a cast including the manifold.

First I would like to say that the intake ports are beautifully laid out with a generous angle between the runner and valve, giving a large inner radius. Much better than some people give them credit for and certainly much, much better than most engines of the time. There are some limitations, that we will discuss later.

The two 101-series cylinder heads have the same part number and the same casting. Port dimensions are, however, different. The 101 “normale” head is limited to simple machining, with especially a restrictive port bowl (see picture), that anyway was good enough for the 80 bhp of the normale. The 101 veloce is further machined (from factory it seems) with a much more open bowl, and valve seat dimensions essentially identical to the later 105 head. Both 101 heads have the ports cast slightly lower, giving a very slight reduction of the port inner radius, compared to the 105 head. The radius is anyway generous and flow between the 101 veloce and the 105 head is almost identical (will show flow figures in a later post).

Note: pictures of the RTV casting and of the 101 normale port were taken after cutting a three-angle (30-46-70 deg) seat. You can clearly see the difference to the 101 Veloce and 105 RTV castings that have the regular 30 deg seats.
Attached Images
     
PSk and horsewidower like this.
DonMartin is offline  
Sponsored Links
Advertisement
 
post #34 of 56 (permalink) Old 07-24-2018, 06:16 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Ideal port dimensions?

Ports that are bigger will flow more, but, if we want any decent low-mid speed performance as well, then we must very carefully keep the minimum port area at the correct dimension and also position. As an example did Alfa themselves reduce the inlet port of the 1600 engine from 32 mm to 29 mm to improve performance.

In the figure I have included the effective cross section areas for the three 1300 heads and also my interpretation of Kartamalakis and Vizards respective recommendations (valve guide and stem areas are accounted for). It also includes my take on ideal section areas, which unfortunately cannot be achieved without port filling, and I have therefore also included a proposal of more practical section areas where I stick to the original port-runner dimensions. My proposals are based on my understanding of how minimal section areas (MSA) should be placed to maximize the pressure recovery from maximized flow speed.

All original cylinder heads and also Kartamalakis suggestion for 1300 (if I understood his book correctly) have the MSA where the valve guide protrudes into the port, around 30 mm from the valve. There are at least two problems with that. The first is that the distance for pressure recovery is very short and the second that guide and valve stem lead to flow disturbance and turbulence (exactly at MSA) which effectively limits the peak flow speed. Modern four-valve engines have their MSA 100 mm or more from the valve, and my take, with the limitations of the 1300 ports and runners, is that we can in best case have the MSA around 60 mm from the valve in a relatively turbulence free area and from there open up towards the valve seat to recover pressure.

The figure shows that the 101 heads have slightly bigger port area before the valve guide than the 105 head (29.8 mm mean diameter versus 29.2 mm) but there is quite some variation from cylinder to cylinder, especially for the 101. I don’t claim that my cylinder heads are representative - maybe there are 101 heads around with smaller ports/runners? My “ideal” proposal suggest a MSA of 28.2 mm, but we are not allowed to add material.

The 101 normale is very narrow at the valve guide and especially in the bowl where the mean diameter is only 30.0 mm versus 32.3 mm for the 105 head. The valve seat inner diameter is 31.5, 33.0 and 32.8 for the 101, 101 Veloce and the 105 respectively.

As you notice, my “practical” proposal is based on keeping the ports/runners unmodified and heavily rework the ports by opening up substantially around the valve guide and in the bowl and seat areas. The final 10 mm is even more aggressively opened up than the proposals from Kartamalakis and Vizard. This particular area is important to slow down flow enough before it turns at the valve opening. This requires a very open valve seat area blended into the bowl, probably without any extended 30 to 45 and 70 deg three-angel cuts but rather going directly from 30 to a short 70 deg cut. For the 1300 I suggest 35.3 mm seat inner diameter, 34.4 mm where the seat ring starts and 33.7 mm 10 mm from the valve. The “Ideal” proposal has 36.2, 34.8 and 33.1 mm, but that would only leave a 0.5 mm wide valve seat, only suitable for bespoke racing cars.

With that said I want to point out that both Kartamalakis and Vizard have probably forgot more engine builds than I have ever started. So, if you follow my ideas it is on your own risk. For now the “practical” proposal is only theories that I would like to try out in the flow bench.
Attached Images
 
DonMartin is offline  
post #35 of 56 (permalink) Old 07-24-2018, 03:07 PM
PSk
Registered User
 
Join Date: May 2003
Location: Tauranga New Zealand
Posts: 11,304
Send a message via AIM to PSk Send a message via Yahoo to PSk
Could you run a skinnier valve stem? Valve looks short and fat ... compared to modern engines.
Pete

'71 1750 Series 2 GTV:
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
156 Series 1 v6 ... and remember it's all just opinions
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

Last edited by PSk; 07-24-2018 at 08:19 PM.
PSk is offline  
post #36 of 56 (permalink) Old 07-24-2018, 07:11 PM
Registered User
 
Join Date: May 2003
Posts: 1,952
good stuff... subscribed
davbert is offline  
post #37 of 56 (permalink) Old 07-25-2018, 05:37 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Quote:
Originally Posted by PSk View Post
Could you run a skinnier valve stem? Valve looks short and fat ... compared to modern engines.
Pete
Fully agree. Acrtually tried with 7 mm stem valve (9 mm original) but it gave just a small improvement. Will show the result later. Anyway, I need to stick with the original valves for FIA K.
DonMartin is offline  
post #38 of 56 (permalink) Old 07-25-2018, 05:45 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Flow testing: Standard cylinder heads

Right, enough theory! Let’s have some real testing.

The first test was to compare the three 1300 standard heads. The 101 normale head flow least (64 cfm) due to the restricted bowl and narrow passage at the valve guide. The 101 Veloce and 105 heads are very similar and the little advantage for the 101 Veloce head comes from the slightly wider seat inner diameter. Both these heads are also clearly flow limited above 8-9 mm valve lift (73 cfm) by the narrow valve guide passage. A comparison was also done with the 2000 engine flow reported by Kartamalakis. The 1300 head flow around 73 cfm, which is similar to an exhaust port of a 2000 (not in figure) and substantially less than the much bigger 2000 inlet port at around 89 cfm.
Attached Images
  
DonMartin is offline  
post #39 of 56 (permalink) Old 07-25-2018, 06:10 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Valve guide modification

The modifications are primarily done on the 105 head since I can live without it in case I overdo the mods. Once I am happy with the mods, these will be implemented in the 101 head, which I intend to use in the FIA K engine. The 101 Veloce head will stay original.

I test modification in small steps to see what different modifications give, rather than just go for my theoretical best and learn nothing.

Since all 1300 heads were flow limited from the port, the first modification was to see the influence of modifying the valve guide. This is a standard modification and following common practice, I reduced the length 2 mm and tapered the protruding part. This very simple modification increased flow almost 3 cfm to 75.6 cfm at 12 mm valve lift and removed the port limitation (so far).
Attached Images
 
DonMartin is offline  
post #40 of 56 (permalink) Old 07-25-2018, 06:43 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Valve seat angles

The second test was to see the influence from valve seat angle cuts. According to Vizard the standard 30 deg seat angle is actually more efficient than the common modification to a 45 deg seat. I will anyway keep the standard valves and seat angle so this is good news.

Vizard, and many others, recommend three-angle cuts, 30-45-70 deg while RJ claims that it is even better to just have a 30-70 deg combination where the 70 deg cut is taken all the way out to the 30 deg inner seat diameter. Both these modifications where tested and compared to the standard 30 deg, with its sharp transition to the almost 90 deg port.

To save enough seat material for final design the inner diameter of the seat was only increased to 34.0 mm (from 32.8 mm) with a 30-46-70 deg cut. Finally, the 70 deg cutter was used to remove the 46 deg cut and go to the 30 deg inner seat diameter of 34.0 mm. Once I have decided for the final valve seat shape I will exploit the full potential for 35.3 mm seat inner diameter giving a seating width of around 1.0 mm.

The results show that increasing the inner seat diameter from 32.8 mm to 34.0 mm with either the 3-angle or 2-angle cut give a massive (almost 20% ) improvement in flow at mid valve lift. The three-angle cut is best at lifts below 6 mm with little difference below 3 mm, while the 30-70 deg combo is best from 6 mm lift and higher, which is probably more important. The flow increased as much as to hit the new port limit with the modified guide. If the port is further properly modified the valve seat seems to be good for at least 80 cfm. Still we haven’t touched the port itself at all yet.

I tried a three angle cut on the 101 head too. And also here there was no improvement of peak flow but a massive improvement at lifts bellow 8 mm. I makes sense that no one recommends 12 mm lift camshafts for 101 motors. But, we are far from done yet!

It is very clear that a good valve seat job cannot help peak flow if the port or bowl is flow limited.
Attached Images
     
PSk and horsewidower like this.
DonMartin is offline  
post #41 of 56 (permalink) Old 07-25-2018, 10:38 AM
Registered User
 
Join Date: May 2003
Posts: 1,952
are the 1600 and 1750 heads closer to the 1300 in port shape and dimensions. while the 2000 heads a separate design all to together?
davbert is offline  
post #42 of 56 (permalink) Old 07-25-2018, 11:10 AM
Moderator
Platinum SubscriberSuper Moderator
 
Gordon Raymond's Avatar
 
Join Date: Jul 2006
Location: Northfield, Illinois
Posts: 10,993
I've been reading and noted some parallel discoveries. GTA piston dome shape. Flat topped worked better than round. Ron Neal at Ausca pre-flow bench experimented with valve seat angles, port shape and funnel seats. I use titanium valves with 7 mm stems.
Pictures below. This was the by-guess-and-by-gosh dyno test method.
Attached Images
   
DonMartin likes this.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
Gordon Raymond
Illinois SNO Alfa Chapter Director
Illinois SNO Ferrari Chapter Director
and sometimes, CONFUSED AND INCORRECT, but Larry helps me out.
Now:
Illinois Chapter Director AND Charter Member of THE CONFUSED AND INCORRECT
Gordon Raymond is offline  
post #43 of 56 (permalink) Old 07-25-2018, 04:56 PM
PSk
Registered User
 
Join Date: May 2003
Location: Tauranga New Zealand
Posts: 11,304
Send a message via AIM to PSk Send a message via Yahoo to PSk
Curious as to why we don't just cut the valve guide and shape it to the port wall shape. This is what my engine builder 30 years ago did to my Sud race engine ... do we really care on a race engine if a guide wears out a little tiny (and it would be tiny) bit faster?

Or does the taper on the guide help the airflow around the valve stem?
Pete

'71 1750 Series 2 GTV:
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
156 Series 1 v6 ... and remember it's all just opinions
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
PSk is offline  
post #44 of 56 (permalink) Old 07-26-2018, 04:19 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Things that work and things that don't

Further valve guide modification:
To relieve the port flow limitation further, a new valve guide was prepared and installed 2 mm further out (total 4 mm less protrution) and shaped to guide flow around. This modification had no practical influence. It could be related with that the valve guide actually directs part of the flow to the short radius which is traded for the slightly increase in section area.

Valve shape
An often recommended modification is to remove the lip above the seat of the original valve. A std valve was modified and compared to a non-modified valve. A third valve was modified even further to see if a flatter back side would help flow. Neither of these changes made any noticeable difference to flow according to my measurements. The flow is however limited by the port at lifts above 9 mm, so it is unclear if an improvement would be visible at higher lifts.

Undercut valve stem
The 101 and 105 series, 1300 alfa engines have valves with quite large 9 mm valve stems. Earlier 750 series 1300 alfa engines and the later Twin Spark engines, just as many other manufacturers, use valve with the more narrow 8 mm valve stems. Aftermarket valve manufacturers claim that more narrow valve stems improve flow, which seems natural. To simulate 8 mm stems, undercut to 7 mm outside the valve guide, the flat back valve from the earlier flow tests was grinded. The flow tests show some improvement (76.5 CFM instead of 75.6 CFM), but it is so small that it is within the error margin.
Gordon Raymond likes this.
DonMartin is offline  
post #45 of 56 (permalink) Old 07-26-2018, 05:14 AM Thread Starter
Registered User
 
Join Date: Mar 2010
Location: Lund Sweden
Posts: 205
Opening up port and bowl and biasing the port

Finally I started opening up the port and bowl area.

From the previous tests we see that the ports are again limiting flow. The throat area is still not touched at all, and much smaller than I think is relevant. To see if flow can pick-up further, and guide if further porting is relevant I opened up the port roof and side just at the valve guide. This was only done on the “front” side (in engine direction) of the valve guide to bias the port. The reason to bias the port is to hopefully introduce some swirl that might speed up combustion slightly. I don’t know if that will work though.

The flow at high valve lifts did indeed pick-up dramatically, and peak flow at 12 mm is now 90.4 CFM. The head is now actually flowing better than the standard 2-liter with 44 mm valves! I still have a little bit of grinding to do, to reach the “practical” dimensions I suggested in a previous post. Once I install the final valve guides and cut the final seats I will open up to those dimensions and flow maybe will improve a little bit more.

Hopefully you can see from the pictures of the port, how it is biased and curves around the guide (left side). It involves quite some grinding in this area! The original casting marks are present in the bottom port area before the valve guide (head is upside down). Actually I would like to reduce this area to 28.2 mm diameter, but I have to stick with the original dimension of about 29 mm, since FIA K does not allow adding any material.
Attached Images
   
horsewidower likes this.
DonMartin is offline  
Reply

Quick Reply
Message:
Options

Register Now



In order to be able to post messages on the Alfa Romeo Bulletin Board & Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.

User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.

Password:


Confirm Password:
Email Address
Please enter a valid email address for yourself.

Email Address:
OR

Log-in










Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page
Display Modes
Linear Mode Linear Mode



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

 
For the best viewing experience please update your browser to Google Chrome