28.01.2015 Views

the performance and metallurgical efficiency of a snif sheer ... - Pyrotek

the performance and metallurgical efficiency of a snif sheer ... - Pyrotek

the performance and metallurgical efficiency of a snif sheer ... - Pyrotek

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

THE PERFORMANCE AND METALLURGICAL EFFICIENCY<br />

OF A SNIF SHEER P-30HB DEGASSING UNIT<br />

Presented at <strong>the</strong> International Melt Quality Workshop,<br />

Madrid, Spain, 25-26 October 2001<br />

Neil J Keegan 1 , Robert A Frank 2 , Nicholas Towsey 3 , Angela Hardman 4 ,<br />

1 <strong>Pyrotek</strong> Engineering Materials Ltd., Dudley, West Midl<strong>and</strong>s, UK.<br />

2<br />

<strong>Pyrotek</strong> Inc., SNIF Systems, 765 Old Saw Mill River Road, Tarrytown, NY, USA<br />

3 VAW aluminium AG, Research <strong>and</strong> Development, Bonn, Germany.<br />

4<br />

London & Sc<strong>and</strong>inavian Metallurgical Co. Limited, Ro<strong>the</strong>rham, South Yorkshire, UK.<br />

Abstract<br />

The test facility at VAW’s Rheinwerk Centre 40 <strong>of</strong>fers a<br />

unique opportunity to quantify <strong>and</strong> baseline <strong>the</strong><br />

<strong>metallurgical</strong> <strong>efficiency</strong> (i.e. hydrogen <strong>and</strong> inclusion<br />

removal <strong>efficiency</strong>) <strong>of</strong> a SNIF degassing system under<br />

controlled conditions. Started early in 2001, this<br />

programme <strong>of</strong> work is <strong>the</strong> 3 rd phase in <strong>the</strong> joint partnership<br />

on metal treatment <strong>of</strong> wrought aluminium alloys,<br />

augmenting <strong>the</strong> earlier filtration programme carried out<br />

between 1995 <strong>and</strong> 2000.<br />

This paper describes <strong>the</strong> major steps proposed in this 3 rd<br />

phase <strong>of</strong> work, along with some preliminary results. These<br />

include <strong>the</strong> work carried out to date to assess <strong>the</strong> best<br />

technique to evaluate inclusion removal. This primarily<br />

focuses on <strong>the</strong> LiMCA method’s difficulty in measuring<br />

accurately inclusion concentration after a degasser.<br />

PoDFA <strong>and</strong> Prefil were used to supplement this study.<br />

This paper also presents <strong>the</strong> preliminary results <strong>of</strong> <strong>the</strong><br />

initial baseline evaluation <strong>of</strong> <strong>the</strong> SNIF system using argon<br />

over a range <strong>of</strong> rotor speeds. Once this baseline study has<br />

been completed <strong>the</strong> work will move on to study <strong>the</strong><br />

interaction with grain refiners <strong>and</strong> CFF’s <strong>and</strong> <strong>the</strong> impact <strong>of</strong><br />

Cl 2 on <strong>the</strong> inclusion removal <strong>efficiency</strong>.<br />

Superimposed on <strong>the</strong> parameter variations would be <strong>the</strong><br />

longer term influences <strong>of</strong> real conditions such as<br />

component wear, dross build up, etc. It is hoped that <strong>the</strong><br />

effects <strong>of</strong> each parameter can be isolated from <strong>the</strong>se<br />

influences.<br />

1.0 Introduction<br />

In this research work we have historically adopted a “layer<br />

on layer” approach so that we can develop <strong>and</strong> build upon<br />

<strong>the</strong> results obtained in earlier phases. The first phase <strong>of</strong><br />

our investigation looked at ceramic foam filters (CFF’s) in<br />

isolation <strong>of</strong> o<strong>the</strong>r in-line treatments to give us a<br />

fundamental underst<strong>and</strong>ing on <strong>the</strong> filtration process <strong>and</strong> a<br />

<strong>performance</strong> baseline. In <strong>the</strong> second phase <strong>of</strong> this research<br />

we introduced grain refiner into <strong>the</strong> system that we had<br />

carefully baselined. This phase was intended to be closer<br />

to typical production conditions existing in many DC cast<br />

houses. In this latest work we start to consider <strong>the</strong> impact<br />

<strong>of</strong> a degasser on downstream metal quality. Having<br />

established its <strong>performance</strong> on its own we will <strong>the</strong>n move<br />

on in <strong>the</strong> future to look at <strong>the</strong> combined effects <strong>of</strong> a<br />

degasser <strong>and</strong> a CFF on <strong>the</strong> quality <strong>of</strong> <strong>the</strong> metal processed.<br />

Naturally a most important area to investigate is how <strong>the</strong><br />

<strong>performance</strong> <strong>of</strong> <strong>the</strong> SNIF varies with its life <strong>and</strong> what<br />

impact any build up on <strong>the</strong> walls or <strong>the</strong> degradation <strong>of</strong> <strong>the</strong><br />

rotor etc., has on <strong>the</strong> overall <strong>performance</strong> <strong>of</strong> <strong>the</strong> system.<br />

2.0 Experimental Procedure<br />

The trials in this study were conducted using <strong>the</strong> specially<br />

dedicated, production scale R&D unit at <strong>the</strong> VAW<br />

Rheinwerk plant. AA1050 alloy, batched using reduction<br />

line metal <strong>and</strong> cast into ingots by <strong>the</strong> direct chill process at<br />

a flow rate <strong>of</strong> 10 tonne/hr, was used throughout. Metal<br />

quality measurements were made in <strong>the</strong> launder using<br />

LiMCA <strong>and</strong> PoDFA techniques. The inclusion content <strong>of</strong><br />

<strong>the</strong> AA1050 metal flowing from <strong>the</strong> furnace, was varied<br />

by a stirring or settling practice. To markedly raise <strong>the</strong><br />

inclusion content, <strong>the</strong> melt was air-stirred for five minutes<br />

prior to casting.<br />

Key to <strong>the</strong> success <strong>of</strong> this work is <strong>the</strong> ability to be able to<br />

carry out meaningful inclusion measurements downstream<br />

<strong>of</strong> a degasser. The options available were LiMCA run in<br />

normal mode using extension tubes as recommended by<br />

Bomem, LiMCA pressure mode measurements as<br />

developed by VAW R+D department 1 , Prefil flow rate<br />

curves <strong>and</strong> PoDFA analysis <strong>of</strong> <strong>the</strong> solidified Prefil sample.<br />

The LiMCA has a problem in measuring inclusion<br />

concentration downstream <strong>of</strong> a degasser because <strong>of</strong> <strong>the</strong><br />

presence <strong>of</strong> gas bubbles in <strong>the</strong> metal. These are counted<br />

along with inclusions. During LiMCA testing a small<br />

volume <strong>of</strong> liquid aluminium is drawn through a 300µm<br />

orifice in a glass tube. The LiMCA has electrodes inside


<strong>and</strong> outside <strong>the</strong> tube. The electrical conductivity <strong>of</strong> <strong>the</strong><br />

metal is changed as non-conducting inclusion particles are<br />

drawn into <strong>the</strong> tube along with <strong>the</strong> liquid metal. The<br />

particles can, <strong>the</strong>refore, be counted <strong>and</strong> <strong>the</strong>ir size<br />

estimated in a way similar to that <strong>of</strong> a Coulter counter.<br />

The normal way that <strong>the</strong>se measurements are done is on<br />

<strong>the</strong> up cycle known as <strong>the</strong> normal or vacuum mode. In<br />

order to try to discount <strong>the</strong> effect <strong>of</strong> gas bubbles a method<br />

was developed by VAW R+D 1 whereby <strong>the</strong> <strong>the</strong> counting is<br />

done on <strong>the</strong> down cycle as <strong>the</strong> metal is ejected through this<br />

orifice. This is referred to as <strong>the</strong> pressure mode.<br />

Recently Comalco 2 reported some work which looked at<br />

<strong>the</strong> pressure method <strong>and</strong> <strong>the</strong> use <strong>of</strong> extension tubes<br />

independently <strong>of</strong> our testing <strong>and</strong> whilst <strong>the</strong>y felt it had<br />

potential <strong>the</strong>y raised concerns that inclusion losses were<br />

occurring during <strong>the</strong>se measurements.<br />

In our work our first aim was to verify that <strong>the</strong> use <strong>of</strong> 2<br />

<strong>and</strong> 3 LiMCA’s at different locations with normal mode <strong>of</strong><br />

operation <strong>and</strong> extension tubes gave us repeatable results.<br />

We <strong>the</strong>n wanted to establish what errors we had by using<br />

<strong>the</strong> pressure mode <strong>and</strong> extension tubes. To do this we<br />

carried out trials running <strong>the</strong> LiMCA’s side by side in our<br />

casting line without gas bubbles in <strong>the</strong> system – i.e. before<br />

<strong>the</strong> installation <strong>of</strong> <strong>the</strong> degasser. We <strong>the</strong>n carried out<br />

similar measurements downstream <strong>of</strong> a SNIF degasser on<br />

a production line in <strong>the</strong> Rheinwerk plant to make<br />

comparisons <strong>of</strong> <strong>the</strong> results in <strong>the</strong> presence <strong>of</strong> bubbles. As a<br />

result <strong>of</strong> this we felt that we were in a position to progress<br />

on to install <strong>the</strong> SNIF P30 on <strong>the</strong> experimental casting<br />

line. We also concluded that N20 <strong>and</strong> possibly higher<br />

would now be used as our baseline as N15 was considered<br />

too sensitive to gas bubbles.<br />

Figure 1 shows <strong>the</strong> SNIF SHEER P30-HB as installed on<br />

<strong>the</strong> experimental casting line during its commissioning<br />

cast. This is a sealed single rotor SNIF system. Figure 2<br />

shows a photograph <strong>of</strong> <strong>the</strong> experimental casting unit <strong>and</strong><br />

launder runs taken from <strong>the</strong> ro<strong>of</strong> <strong>of</strong> <strong>the</strong> SNIF P30. Metal<br />

enters <strong>the</strong> SNIF from <strong>the</strong> 20 tonne cylindrical casting<br />

furnace on <strong>the</strong> left before exiting via <strong>the</strong> launder on <strong>the</strong><br />

bottom right h<strong>and</strong> side <strong>of</strong> <strong>the</strong> photograph. This launder<br />

continues fully 9m before reaching <strong>the</strong> casting launder.<br />

The experimental casting line was set up like this in order<br />

to try to limit <strong>the</strong> influence <strong>of</strong> gas bubbles on our LiMCA<br />

measurements. Two LiMCA’s can also be seen in this<br />

photograph as used during <strong>the</strong> experimental programme.<br />

One is just in front <strong>of</strong> <strong>the</strong> SNIF P30 on <strong>the</strong> bottom left<br />

h<strong>and</strong> side <strong>and</strong> <strong>the</strong> second is at <strong>the</strong> top centre <strong>of</strong> <strong>the</strong><br />

photograph showing measurements being conducted<br />

downstream <strong>of</strong> <strong>the</strong> SNIF unit just ahead <strong>of</strong> <strong>the</strong> casting<br />

launder.<br />

Figure 1: SNIF SHEER P30-HB on Rheinwerk casting line.<br />

Figure 2: Experimental Casting Pit Layout.<br />

3.0 Results <strong>and</strong> Discussion<br />

3.1 LiMCA testing using normal <strong>and</strong> extension tubes in<br />

normal <strong>and</strong> pressure mode without argon gas bubbles in<br />

<strong>the</strong> system.<br />

Figure 3 shows <strong>the</strong> LiMCA traces for 3 LiMCA’s ran<br />

side by side with extension tubes on our casting line<br />

without bubbles i.e. without a degasser in <strong>the</strong> line. We can<br />

see that <strong>the</strong> three units are quite repeatable over <strong>the</strong> cast as<br />

a whole. There is a higher degree <strong>of</strong> scatter which is quite<br />

normal when running with extension tubes. The areas <strong>of</strong><br />

<strong>the</strong> traces outlined with <strong>the</strong> red circles show where one <strong>of</strong><br />

<strong>the</strong> units was run in pressure mode. The first <strong>and</strong> last is<br />

where <strong>the</strong> LiMCA with <strong>the</strong> black line is run in pressure<br />

mode. In <strong>the</strong> first instance <strong>the</strong> pressure mode seems to<br />

have made very little difference. In <strong>the</strong> latter case as with<br />

<strong>the</strong> LiMCA shown by <strong>the</strong> pink line <strong>the</strong>re may be an<br />

indication that <strong>the</strong> readings are slightly lower in this mode<br />

than in normal mode. In contrast, however, o<strong>the</strong>r casts<br />

showed <strong>the</strong> pressure mode sometimes gave higher results


than vacuum. This was probably, given general<br />

experiences with extension tubes, due to <strong>the</strong> scatter that<br />

occurs whilst running <strong>the</strong>se tubes.<br />

that <strong>the</strong>re is a clear <strong>of</strong>fset between <strong>the</strong> counts obtained<br />

with extension tubes as compared with those gained using<br />

<strong>the</strong> normal tube. After approximately 25 minutes <strong>the</strong><br />

second LiMCA shown with <strong>the</strong> blue line was run in<br />

pressure mode. In this case, admittedly at low inclusion<br />

levels, <strong>the</strong> pink <strong>and</strong> blue lines show little difference<br />

between <strong>the</strong> pressure <strong>and</strong> normal mode.<br />

Figure 3: 3 LiMCA’s run in normal mode with extension<br />

tubes - no argon gas bubbles – i.e. without degasser.<br />

Figure 4 shows a trial where all three units were run<br />

initially in normal mode with normal tubes for <strong>the</strong> without<br />

bubbles case i.e. without degasser. The repeatability is<br />

again very good. After 35 minutes an extension tube was<br />

put on <strong>the</strong> second LiMCA, <strong>the</strong> one represented by <strong>the</strong> blue<br />

line. It can be seen that this has led to a lower level <strong>of</strong><br />

inclusions being counted.<br />

Figure 5: 3 LiMCA units, 1run in normal mode with a<br />

normal tube, 2 with extension tubes – no argon gas bubbles.<br />

3.2 LiMCA testing using normal <strong>and</strong> extension tubes in<br />

normal <strong>and</strong> pressure mode with argon gas bubbles in <strong>the</strong><br />

system.<br />

Having carried out some validation trials using <strong>the</strong><br />

LiMCA without bubbles, we <strong>the</strong>n went on to evaluate<br />

some production metal downstream <strong>of</strong> a SNIF degassing<br />

unit which represented <strong>the</strong> ‘with argon bubbles’ case.<br />

Figure 4: 3 LiMCA units run in normal mode, 2 with<br />

normal tubes, 1 with extension tube – no argon gas bubbles.<br />

Figure 5 shows clearly <strong>the</strong> <strong>of</strong>fset that occurs when using<br />

extension tubes compared with normal tubes. This is again<br />

for <strong>the</strong> without bubbles case. The green line is our<br />

reference line <strong>and</strong> shows <strong>the</strong> trace for normal mode with<br />

normal tubes. The blue <strong>and</strong> pink lines <strong>of</strong> <strong>the</strong> second <strong>and</strong><br />

third LiMCA are run using extension tubes. It can be seen<br />

Figure 6: LiMCA results before <strong>and</strong> after a SNIF R140 on<br />

Rheinewerk production unit, with extension tubes.<br />

Figure 6 shows a typical trial result. The pink line is from<br />

<strong>the</strong> LiMCA downstream <strong>and</strong> close to <strong>the</strong> degassing unit.<br />

It can be seen that <strong>the</strong> first measurements made in normal<br />

mode using extension tubes show that <strong>the</strong> LiMCA counts<br />

are much higher after <strong>the</strong> degasser than before. This is


elieved to be as a result <strong>of</strong> counting bubbles. When we<br />

switched <strong>the</strong> LiMCA to pressure mode we can see that for<br />

<strong>the</strong>se conditions we appear to be eliminating most <strong>of</strong> <strong>the</strong><br />

gas bubbles from our counts. We do not know <strong>the</strong> loss <strong>of</strong><br />

inclusions due to <strong>the</strong> pressure mode <strong>and</strong> extension tube<br />

combined. When we switched back to normal mode <strong>the</strong><br />

line immediately returned to a level similar to that<br />

detected before. Going back to pressure mode again gave<br />

us a value similar or slightly lower than that <strong>of</strong> <strong>the</strong><br />

incoming metal counts.<br />

Figure 8: LiMCA histogram results for after <strong>the</strong> SNIF<br />

SHEER P30-HB with extension tubes in pressure <strong>and</strong><br />

vacuum mode.<br />

Figure 7: LiMCA results before <strong>and</strong> after <strong>the</strong> SNIF<br />

SHEER P30-HB with extension tubes in pressure <strong>and</strong><br />

vacuum mode.<br />

Figure 7 shows one <strong>of</strong> <strong>the</strong> first LiMCA trials run with <strong>the</strong><br />

SNIF P30 for <strong>the</strong> “with gas bubbles” scenario. In this trial<br />

<strong>the</strong> SNIF was set to run at 500 rpm <strong>and</strong> with 3Nm 3 /hr<br />

argon. We were now varying <strong>the</strong> SNIF parameters <strong>and</strong><br />

along with <strong>the</strong> PoDFA trying to establish <strong>the</strong> measurement<br />

errors for each setting due to <strong>the</strong> different inclusion <strong>and</strong><br />

bubble losses. It must be assumed that <strong>the</strong> inclusion <strong>and</strong><br />

gas bubble size distribution is also changing accordingly.<br />

The before LiMCA curve is in pink <strong>and</strong> it can be seen that<br />

<strong>the</strong> incoming metal was very clean. This was verified by<br />

<strong>the</strong> PoDFA counts taken to supplement <strong>the</strong> LiMCA<br />

measurements. During this trial consecutive points were<br />

taken in normal <strong>and</strong> pressure mode using extension tubes.<br />

The pressure mode has lead to lower counts than was <strong>the</strong><br />

case for <strong>the</strong> normal mode but <strong>the</strong> downstream LiMCA<br />

values can be seen to be still higher than those <strong>of</strong> <strong>the</strong><br />

upstream LiMCA . The PoDFA showed o<strong>the</strong>rwise being<br />

very clean at less than 0.01mm 2 /kg. It appeared, <strong>the</strong>refore,<br />

that despite having <strong>the</strong> LiMCA over 9m away we were<br />

still counting very small gas bubbles after <strong>the</strong> degasser.<br />

The marked reduction in counts detected using <strong>the</strong><br />

pressure mode as opposed to <strong>the</strong> normal mode is shown<br />

more clearly in Figure 8, <strong>the</strong> histogram for <strong>the</strong> previous<br />

trial.<br />

The SNIF P30 operating parameters were <strong>the</strong>n changed to<br />

check our measurement technique over a wider range <strong>of</strong><br />

settings. In Figure 9 we see that for <strong>the</strong>se parameters <strong>the</strong><br />

LiMCA’s perform differently to before. The incoming<br />

metal is dirtier than for <strong>the</strong> previous trial <strong>and</strong> is shown<br />

with <strong>the</strong> blue line. This was again found to agree with <strong>the</strong><br />

subsequent PoDFA measurements. The downstream<br />

LiMCA SNIF values are again shown in pink. This time<br />

when we switch to pressure mode as represented by <strong>the</strong><br />

red points, we seem to be excluding most <strong>of</strong> <strong>the</strong> gas<br />

bubbles generated for <strong>the</strong>se SNIF settings. The PoDFA’s<br />

were again as clean as for <strong>the</strong> last example which<br />

supported this conclusion.<br />

Figure 9: LiMCA results for before <strong>and</strong> after <strong>the</strong> SNIF<br />

SHEER P30-HB operating at 400rpm with 2Nm 3 /hr argon<br />

run with extension tubes in pressure <strong>and</strong> vacuum mode


Figure 10: Typical PoDFA fields for before <strong>and</strong> after <strong>the</strong><br />

SNIF SHEER P30-HB on <strong>the</strong> experimental casting line for<br />

cast 020401-2<br />

tested this new approach. The downstream SNIF values<br />

are again in pink <strong>and</strong> can be considered to be our<br />

background due to gas bubbles which should not vary for<br />

<strong>the</strong>se particular SNIF settings. We <strong>the</strong>n stirred <strong>the</strong> furnace<br />

during <strong>the</strong> cast. It is assumed our background for<br />

semiquantitative purposes will not alter too much as a<br />

result <strong>of</strong> <strong>the</strong> varying error in inclusion counts, whilst <strong>the</strong><br />

incoming inclusion loading should naturally go up. We<br />

can, <strong>the</strong>refore, judge semi-quantitatively how <strong>the</strong> SNIF<br />

has performed by <strong>the</strong> response <strong>of</strong> <strong>the</strong> downstream line. If<br />

as in this case <strong>the</strong> line remains fairly constant despite a<br />

large increase in loading upstream <strong>of</strong> <strong>the</strong> SNIF it can be<br />

assumed that <strong>the</strong> SNIF is doing a good job <strong>of</strong> cleaning <strong>the</strong><br />

metal.<br />

Figure 10 shows some typical fields from <strong>the</strong><br />

<strong>metallurgical</strong> evaluation <strong>of</strong> <strong>the</strong> Prefil samples. The left<br />

h<strong>and</strong> side shows <strong>the</strong> before SNIF sample with a large<br />

aluminium carbide layer adjacent to <strong>the</strong> top <strong>of</strong> <strong>the</strong> sample.<br />

This has a PoDFA count <strong>of</strong> 0.6 mm2/kg. The after SNIF<br />

sample can be seen to be very much cleaner with much<br />

less <strong>of</strong> a layer being formed in <strong>the</strong> sample. This had a<br />

PoDFA count <strong>of</strong> 0.2mm2/kg.<br />

Figure 12: LiMCA results during a trial using <strong>the</strong><br />

“Background method” for <strong>the</strong> SNIF SHEER P30-HB on<br />

<strong>the</strong> experimental casting line.<br />

If this is <strong>the</strong> case as is shown again in Figure 13 we have<br />

what we will call <strong>the</strong> best case scenario. The downstream<br />

line has remained constant so we have what we will grade<br />

as an “A” <strong>performance</strong>.<br />

Figure 11: PoDFA total inclusion contents for <strong>the</strong> SNIF<br />

SHEER P30-HB on <strong>the</strong> experimental casting line.<br />

Figure 11 shows <strong>the</strong> results <strong>of</strong> <strong>the</strong> PoDFA counts for <strong>the</strong><br />

before <strong>and</strong> after SNIF P30 samples taken during our trials.<br />

It can be seen that for <strong>the</strong> clean SNIF unit a good inclusion<br />

removal is consistently being obtained. This shows that<br />

<strong>the</strong> LiMCA is indeed measuring bubbles after <strong>the</strong> SNIF<br />

ra<strong>the</strong>r than inclusions.<br />

It appeared, <strong>the</strong>refore, that we still needed to adapt our<br />

LiMCA measurement technique if we were to be able to<br />

use <strong>the</strong>m successfully downstream <strong>of</strong> a degassing unit.<br />

With this in mind we have come up with a special<br />

approach. We have called this <strong>the</strong> “background method”.<br />

The logic is that we collect data upstream <strong>and</strong> downstream<br />

<strong>of</strong> <strong>the</strong> degasser for a period until <strong>the</strong> readings are<br />

considered to be stable. Figure 12 shows a trial where we<br />

Figure 13: LiMCA results during cast 020501-1 using <strong>the</strong><br />

“Background method” for <strong>the</strong> SNIF SHEER P30-HB.


In Figure 14 where we changed <strong>the</strong> SNIF P30 parameters<br />

to a more unsuitable setting we see <strong>the</strong> downstream line<br />

has increased along with <strong>the</strong> upstream line after stirring.<br />

This is what we rank as a “B” <strong>performance</strong>. Similarly if<br />

<strong>the</strong> downstream line appears <strong>the</strong> same or worse than <strong>the</strong><br />

before SNIF curve <strong>the</strong>n we shall have what we consider to<br />

be a “C” <strong>performance</strong>.<br />

is working fundamentally as expected <strong>and</strong> appears in line<br />

with <strong>the</strong> positive inclusion removal data obtained. This<br />

obviously needs to be verified over a much larger number<br />

<strong>of</strong> trials before a proper comparison with <strong>the</strong> inclusion<br />

removal data can be made.<br />

Figure 16: Alscan hydrogen removal efficiencies for <strong>the</strong><br />

SNIF SHEER P30-HB on <strong>the</strong> experimental casting line.<br />

Figure 14: LiMCA results during cast 030501-2 using <strong>the</strong><br />

“Background method” for <strong>the</strong> SNIF SHEER P30-HB.<br />

These rankings will be validated over a large number <strong>of</strong><br />

casts <strong>and</strong> in conjunction with PoDFA analysis. With this<br />

methodology we aim to make sound assessments <strong>of</strong> <strong>the</strong><br />

inclusion removal capability <strong>of</strong> <strong>the</strong> SNIF <strong>and</strong> how this<br />

changes with varying operational parameters. Due to <strong>the</strong><br />

shortcoming <strong>of</strong> <strong>the</strong> LiMCA in <strong>the</strong> presence <strong>of</strong> gas bubbles<br />

we cannot produce quantitative results. However, with full<br />

PoDFA back up we hope to get a semi-quantitative<br />

classification.<br />

4.0 Conclusions<br />

(1) The “Background” method appears promising but<br />

needs be confirmed.<br />

(2) This <strong>and</strong> PoDFA indicates good inclusion<br />

removal with argon <strong>and</strong> a clean SNIF.<br />

(3) The pressure mode alone is not accurate enough<br />

to compare parameters.<br />

(4) Prefil curves alone appear invalid so it is<br />

necessary to rely on PoDFA metallographic<br />

analysis.<br />

(5) Alscan will be required to supplement inclusion<br />

removal measurements.<br />

5.0 References<br />

(1)<br />

H-P. Krug, & W.Schneider,.<br />

"A contribution to inclusion measurement after in—line<br />

degassers with PoDFA <strong>and</strong> LiMCA", Light Metals 1998,<br />

pp.863-870.<br />

Figure 15: PoDFA total inclusion contents for <strong>the</strong> SNIF<br />

SHEER P30-HB on <strong>the</strong> experimental casting line.<br />

(2) M.Cooksey, T.Ware & M.J.Couper<br />

”Effect <strong>of</strong> pressure cycle <strong>and</strong> extension probe on LiMCA<br />

measurement <strong>of</strong> inclusions”, Light Metals 2001, pp 965-971<br />

Finally, just to confirm <strong>the</strong> inclusion removal that we<br />

anticipate to have occurred during our background method<br />

trials, Figure 15 details <strong>the</strong> accompanying PoDFA counts<br />

for our latest series <strong>of</strong> trials. It can be seen that as was <strong>the</strong><br />

case for <strong>the</strong> first trials we are indeed getting good <strong>and</strong><br />

verifiable inclusion removal using <strong>the</strong> clean SNIF.<br />

Figure 16 shows <strong>the</strong> Alscan hydrogen removal<br />

efficiencies for <strong>the</strong> same trials. This shows that <strong>the</strong> SNIF

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!