Final report of the hydro flow symposium

Final report of the hydro flow symposium

Citation for published version (APA):

Technische Hogeschool Eindhoven (THE). Fac. Werktuigbouwkunde (1973). Final report of the hydro flow symposium. Eindhoven University of Technology.

Document status and date: Published: 01/01/1973

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

FINAL REPORT

of the

'HYDRO FLOW' SYMPOSIUM

Organised by the Power Transmission Department

of the Eindhoven University of Technology, The Netherlands; in collaboration with CETOP

INTRODUCTION

The aim of 'Hydro flow':

The aim of the Symposium was to come to both a transfer and exchange of knowledge and experiences on flow and mass flow measurements ~n oilhydraulic systems, and tohand this practical information, based on scientific analyses, to the engineer in industry who is responsable for the choice of measurement devices fordevelopment and testing laboratories.

Werking method during the 'Hydro flow' symposium

In order to achieve the above mentioned aim, the symposium was organised in a discussion form. The Power Transmission Department hence formulated seven discussion items; upon request of several participants, an· eighth subject has been added (see annexed list of discussion subjects).

This was the working method during the symposium:

First, one of the members of the Power Transmission Department gave a short introduetion on the relevant item, after which a mem-ber of the panel took over and stated his opinio~ thus getting the discussion started.

In this final report the apinion of the Power Transmission

Department will be stated,which servedas an introduetion to the items, tagether with a summary of the discussions without however

going into any details.

The panel on 'Hydro flow':

The panel of the 'Hydro flow' symposium consisted of the following persons:

2

-Prof,Dr.Ir. W.M.J. SchlÖsser, Power Transmission Department (Chai r~nan)

Ing. G. Toet, Power Transmission Department, (chief eng~neer of the Power Transmission Laboratory) Ir. H.A. Essers. Power Transmission Department (responsible for

preparing the symposium) Prof.Dr. D.E. Bowns, Bath University (Fluid Power Centre)

J. Lerich, Thermo Systems Inc. (hot wire and hotfilm anemometers) Ing. F.W. Rosteck, Hydrateehuik K.G. (turbine meters)

Dr. H. Voss, Robert Bosch G.m.b.H (expert in the field of selecting meásurement devices for develonment

laboratories)

D.J.M. Smith, National Engineering Laboratory, Glasgow. Miss H. Herschberg, interpreter

Discussion themes for the 'Hydro.flow' symposium:

Theme I:

What selection has to be made for oilhydraulic purposes: Volume flow measurement or mass flow measur'ement?

Theme 2:

Where are flöw (mass flow) meters applied in hydraulics?

Theme 3:

How can the fields of application be divided in v~ew of required properties of the flow transducers, such as:

a. accuracy?

b. viscosity and density dependence? c. range?

d. max~mum permissible pressure and temperature? e. sensitivity to dirt?

f. energy dissipation? g. purchase costs?

h. stability (long term repeatibility)? i. life expectancy?

J· dynamic properties?

Theme 4:

What type of measurement devices are suitable for these fields of application in view of their properties as compared to the required properties agreed upon ~n item 3?

Theme 5:

Which flow transducers are considered to be most suitable for measuring steady state flow (massflow) in oilhydraulic systems,

in view of the requirements as discussed under item 4?

Theme 6:

Which flow transducers are considered to be most suitable for measuring the dynamic flow behaviour in oilhydraulic systems

4

-Theme 7:

Which equipment, incorporated in the flow transducers can be considered for both steady state a1·d dynamic measurements?

The following item was added during the symposium:

Theme 8:

Which flow transducers are suitable for measuring very small flows (leak flow) up toa maximum of 0.075.10-3 m3/s

Theme I

What selection has to be made for oilhydraulic purposes: Volume flow meaHurement or mass flow measurement?

Opinion of the Power Transmission Department

For hydraulics, flow is the relevant ouantity because of the fact that the components (hydropumps and -motors, cylinders and the like) when applied in hydraulic (hydra-statie) systems are based on the displacement principle.

Thus, the number of revolutions of a hydra-motor is prîmarily determined by the flow effered (and not by the mass flow).

The same applies to a hydraulic cilinder; the speed of the piston is likewise determined by the flow offered, because ~n both cases a vacated volume must be filled by liquid, where the density of this liquid - and with that the mass (flow) - is of secondary importance. So, for this reason, flow-measurement should be

selected for hydraulics, in spite of the fact that mass flowoffers the advantage of being a quantity independent of bath local

pressure and temperature in the system. Application of mass flow measuring implies that a long wi th measuring the mass flow,

pressure and temperature must be measured as well .(although this aften applies to flo~ measuring also because of pressure- and temperature dependance of the calibration curves of the flow meters; to this we will comeback later on more extensively),but also that an (accurate) relation between the density of the measured liquid and pressure and temperature must be known in order to be able to determine the flow.

6 -Summary of the discussions:

The general idea was that in oilhydraulics, volume flow does indeed constitute the relevant factor, and thus volume flow has to be measured. The remark was however made that in some cases a 'detour' via mass flow measurement is preferable or even unavoidable:

a. for measuring.quick flow variations (dynamic measurements): Beyond certain variations in speed limits, the existing volume flow transducers are no langer capable to 'follow' and measure these rapid variations. Here, a mass flow measurement us~ng a hot film anemometer will give the desired results.

b. for calibrating flow transducers: It is much easier to measure a mass rather than a volume with high accuracy. For this reason, a mass (flow) measurement is used for calibrating flow trans-doeers whenever a high degree of accuracy is required: Upön a weighing bridge a tank is ulaced ~n which the discharge line of the flow transducer terminates. One then measures the time necessary to let an accurately known mass flow via the flow transducer into the tank (for examule by using two detectors ulaced along the scale of the weighing bridge, which resnectively send a start ar a stou signal to an electrooie clock whenever the uointer uasses over them). By deviding the known mass by the

measured time, the mass flow can be found, from which in turn the volume flow can be calculated when the fluid density (at the flow transducer!) is known, because:

mass flow

=

density x volume flow

~

=

p(G,p) x Qv

In a formula,this can be expressed as follows:

volume flow Qv -..,.---_.... steady state measurements

mass flow

x difficult calibration P(8,p)

----~~ dynamic measurements easy calibration

Theme 2

Where are flow (mass flow) meters a: t-Jlied ~n hydraulics? Opinion of the Power Transmission Department.

In hydraulic~ a first division in the fields of application for flow meters (and transducers in general), ~s a division into messurement systems. and transmissions. By measurement systems we understand such systems as have been designed for messurement at hydraulic components, i.e. generally laboratory conditions, whereas by transmission~ we mean already existing systems, the object of which is the transmission of energy, hence hydrastatic transmissions in general.

The major difference between these two fields of application chiefly lies in the (highly) divergent conditions under which transducers are applied. Whereas in the first case (measurement systems) both the circumstances in the system itself and the external ones can be well conditioned in most instances, this will by far not always be so in the secend case (transmissions). According to required purpose, the messurement systems can be distinguished in systems for studying elementary phenomena ~n

the field of hydraulics and in those for testing hydraulic

components. In the first instance, we have a distinct application of flow meters for pure research, such as e.g. the measurement at prototype displacement pumps- or motors, leading us, VLa loss-separation, to a better insight into and impravement of the construction; or we might study the pump rib or occurring cavitation phenomena. The second case shows an application of flow meters for production tests assessing quantity determination of losses.

In practice, these two domains mainly differ from each other by the degree of required accuracy of the flow meters. The

application of flow meters in transmissions can be divided ~n

8

-in ordertotrace causes of -interferences and suchJand to check correct operation, and in one wher. ..:he flow meter is used for contineously indicating the flow in the system. In the first instance, the flow meter is not permanently present in the system, but is only applied if interferences and such occur to trace the causes thereof which, on account of the aften

occurring external circumstances, we shall call field• measurements. In the second case, the flow meter is actually permanently

present in the system for flow indicating purposes, an example of which is found in the use of the flow meter as a safeguarding element (airplane hydraulics).

In a diagram the application of flow meters ~n hydraulic systems can be shown as follows:

Fields of application

~n hydraulics

Summary of discussions

During the discussions on this subject~ the division into various fields of applications, as proposed by the Power Transmission Department, was accepted as being correct.

Furthermore an attempt was made - as a resumé of subjects I and 2 - to stipulate where volume flow and where mass flow measurements are applied:

volume flow measurement

'

difficult

"

field measurements calibration

/

production tests easy

/

mass flow measurement research

-JO-Theme 3

How can the fields of application b0 divided in v~ew of required properties of the flow transducers, such as:

a. accuracy?

b. viscosity and density dependence? c. range?

d. max~mum permissible pressure and temperature? e. sensitivity to dirt?

f. energy dissipation? g. purchase casts?

h. stability (long term repeatibility)?

~. life expectancy? J• dynamic proporties?

Opinion of the Power Transmission Department

The preceding question already gave us a. division of application-fields for flow meters in hydraulics.

In view of particular applications,the properties of flow meters will have to be adjusted accordingly.

A first division gave us'measurement systems' and 'transmissions'. As we have already seen, one of the major differences in these fields of application lies in the aften strongly divergent circumstances under which flow meters are applied.

In case of 'measurement systems', the (volume) flow meters are applied ~n (mostly) well to very well conditionable circumstances: good pressure- and temperature adjustment, filtration of the

hydraulic liquid in the system,and practically constant external circumstances. Because of this, the requirements as to viscosity-and density dependenee viscosity-and sensitivity to dirt may be of secondary importance.

In a subsequent subdivision of 'measurement systems', we have

made a distinction between 'research' and 'production tests'·, where we have already stated that the difference here is mainly

determined by the degree of required accuracy; when us1.ng flow meters for research purposes accur · .cy wi ll play a predominant role, to which the other propertie0 will be subordinate (by

accurac~ we mean- besides the accuracy of a single: measurement-the repeatibility of a same measurement as well). Thus a lesser long-term repeatibility (or a smaller range) will be acceptable, if accuracy and repeatibility are very good. High purchase costs will, in view of the fact that these are often to be ~onsidered as a one-time occurence, be of lesser importance7 which also goes for a relatively large energy dissipation, as the measurement systems are primarily intended for measuring at hydraulic components and not for the transmission of energy, as 1.n the case of transmissions, where a relatively large energy dissipat-ion caused by the transducer (as a result of a large pressure-drop over the meter) means a relatively considerable loss. When using flow meters for production tests, the requirements as to

accuracy and repeatibility will be less stringent than for research purposes. Here, however, different requirements as to the properties of the flow meters become more dominant.

Thus, in case of production tests for series production, where

the flow meter will be present for a long period in the same measurement system, stability and possibly also sensitivity to dirt as well as its life expectancy will play a part, while energv dissipation can also become important. As yet, viscosity- and density dependenee and purchase costs will not play an important

part. For flow meters in transmissions the situation is often

quite different. These systems, which primarily transmit energy,

can not be so well conditioried in most cases, neither as re~ards pressure and temperature in the system, nor as regards the

external circumstances; filtratien of these systems is mostly

inferior to that of the measurement systems.

Hence,different demands will be made to the properties of flow meters when applied in 'transmissions'.

/ _{- 12}

-'transmissions'into 'field measureme~ts' and 'flow indication'. Especially in the second case, whe flow meters will be present in the same system for a long time~ such properties as favour a prolonged operation without interruption - also under drastically changing conditions - such as small sensitivity to dirt fore few interruptions for maintenance), a good stability (there-fore seldom calibration of the flow meter is necessary) and a long life expectancy will positively influence the choice of flow meter. The purchase casts and especially those resulting from energy dissipation in the system will play an important part as well. In this field of application, however, only very low de-mands will be made as to accuracy and repeatibility, so that

less importance will be attached to the viscosity- and density dependenee of a flow meter.

In the first case('field measurements'), somedemand will be made as to accuracy and repeatibility (see further on), as aresult of which small viscosity- and density dependenee will be of relatively great importance. Characteristics such as good

stability, small sensitivity to dirt also play a dominant role, whereas a large range, in which the abovementioned demands for accuracy and repeatibility are met with, is also of great

importance since an exchange of transducers during measurements is undesired or even excluded. Especially in this field, the purchase casts will play a very important, if nat predominant role in the choise of flow meter.

As may be clear from the above , the most important

characteristics for a flow meter are accuracy and repeatibility commensurate with a number of other characteristics, such as range, viscosity- and density dependence.

Now,what demands can be made as to accuracy and repeatibility for the various fields of application?

To achieve uniformity in this respect, a proposal for an ISO-Standard has been made by ISO/TC131/SC8 (United Kingdom-2) 8,

which we will adopt with some slight alterations. Unlike the British proposal, we do notintend 1o conneet the maximal tolerabie deviation with the reading, but with the full scale deflection.

For flow, the following applies:(for example fora measuring range of 1 :5)

Maximum limit of systematic error Class A Class B Class C

flow % + 0 ,I + 0.3 .!_0,5

In ·the British proposal it has been forthermore indicated that: a. for class A, the flow meter befare and after the test must be

calibrated with the medium to be measured at temperatures and pressores which will be established during the test. b. for class B, the flow meter must be calibrated immediateil

befare the test, in a manner as described sub a).

c. for class C, proof must be available of a calibration done less than twelve months ago with a medium having physical properties identical to those of the medium to be tested.

The three classes A, B and C now correspond to the fields of application, resp. called 'Research', 'Production tests' and

'Field-measurements'.

We suggest the following maximum va lues for repeatibility: Class A ( 'reseàrch' ) : 0,1%

Class B ('production tests'): 0,25%

Class

c

( 1

field measurements' ): o.s%

Depending on application, usually a larger deviation from the measured value and repeatibility is allowed for 'flow indication 1

(maximum up to abt .

.:!:.

5%, resp. I%).

As to the maximum allowable pressure, we can distinguish two fields in hydraulics:

the field of 'low' pressure the field of 'high' pressure

up to abt. 40 to 50 bar up to abt. 400 bar

14

-As to maximum allowable t,emperature, we can state that the maximal werking temperature for a hy· :aulic liquid does not, in

most cases, exceed 80° C which is due to the fact that over and ·above this temperature the lifespan of the hydraulic liquids quickly shortens, a.o. as a result of oxydation, though in extreme cases werking temperatures of up to 120°C are permissible.

Whether a flow meter (apart from being used for measuring steady state flows, that is to say flows in which depends on the

(around an average value) occur which depends on ·the

fact whether the flow meter confarms to the demands as described above\ can also be used for measuring changing flows and if so, to which frequencies, depends on the dynamic properties of the flow meter in question.

In drafting criteria for descrihing dynamic properties of flow meters, we must distinguish between first-order systems and secend-order systems, that is to say between flow meters the dynamic behaviour of which can be described with the aid of a first-order differential equation (e.g. a turbine meter), resp. with a secend-order differential equation (e.g. the drag force meter):

a. the first-order system:

b. the second-erder system:

bere the dynámic properties can

'

be described by the time constant T

here the dynamic properties are characterised by natural-frequency

we and the damping factor t,; .•

The influence of these factors on the dynamic behaviour is clearly visible from the frequency response characteristics of both systems, which is the most important souree of information for dynamic

properties (see added figure on page 16 ).

Now₁where are flow meters applied for dynamic purposes?

Generally speaking1we can divide flow meters for measuring changing

(research), such as the already mentioned pump rib and cavitation phenomena, and in these for testing dynamic behaviour of

separate components (e.g. production tests of valves) or of complete systems (e.g. in 'field measurements).

In case of 'flow indication' more often than not good dynamic properties of the applied transducers are not appreciated, since the flow meter may not react to short-1ived major flow changes, such as these may for example occur during switching-on of a system. nor to 'quick' periodical changes, such as the pump rib.

IHI

r

q1 ~()~ o• _?() v.J7:

-3o·

1

4H

-60° -9017

first order system

P,~

I

Hl

qo;' q'?'t~;' ~O()

...

o•

1-M

4H

-;'..;>t)o - ~tf()D Fresuency Response 0,~

r=o.-1

r=

q~

r

=

~?)

J::

2,() ~p

second order system

0\

."..v '

Summary of discussions

a. Accuracy: ~n other words, what gr Jes of accuracy do we aim at? The Power Transmission Department therefore put forward the proposal as given on page 13.

To clarify some points regarding a definition of 'acc~racy' as given ~n the above mentioned proposal, a further explanation was g~ven by way of an example (see fig. 1):

I

I

I

"8,-/·~/~~ ,Pro_"oosa( --t---~--11 fig_._l

For class A₁the max~mum systematic error may not exceed ~ 0,5%. This implies that for a range (A:B) of I :5 the systematic error

of the final value amounts to + 0.1%.

The reason for this deviation from the British proposal is that the Power Transmission Department thinks that a figure of~ 0.5% over the total working area in most cases is asking too much or even for the impossible, as an exchange of transducers during the measurements is undesirable.

This is contradictory to Mr. Smith's apinion (N.E.L.), who defended the British proposal, and according to whom indeed more than one transducer has to be used to eventually meet with specifications. The discussion on this point, however, was limited (because this will be further dealt with in ISO-TC 131, leading to a final

- 18

-During the symposium, the proposal of the Power Transmission Department was used for practical ourooses.

b. Viscosity-and density dependence: here, the role of the following two points was explained.

1. the dependenee of a transducer's calibration curve on viscosity and density;

2. to which extent can viscosity and density he kept constant, so as to limit their influence to a minimum.

c. Range: this has already been discussed comprehensively under point a) (accuracy), because required accuracy and range are closely related to each other, as also shown in fig. 1. Range was therefore defined as:

Range working area in relation to the maximum allowable inaccuracy or else: (see fig. 1) range = A:B

where B

=

upper working area limit;mostly determined by the transducer's construction.

A lower range limit; mostly determined by the max1.mum allowable inaccuracy.

Furthermore, we'd like to state that the larger the range 1.s the better this is for the choice of transducer.

Another aspect brought up in the discussion covered the ranges for several transducers in one series:

The need for Renard numbers for a series of flow transducers (i.e. ranges for these transducers are arranged according to Renard

numbers) was not considered to he so evident, as it 1.s for pump series, because of the fact that the flow affered to the trans-ducer is determined bath by the swept volume and the number of pump revolutions;an increase in swept volume is translated by a decrease in the maximum number of revolutions.

d. Maximum admissible pressure and temperature:

T':le Power Transmission Department's proposal was accepted

subject to an alteration intheupper limit of the 'high' pressure region which was considered as being too low. It was increased to 500 and in extreme cases even to 600 bar, giving us the following final division:

pressure 'low' pressure region: up to abt.

_-

+ 40 - 50 bar 'high' pressure region: up to abt. 500 600 temperature normal systems: 80°C

extreme cases : 120°C.

e. Sensitivity to dirt: 'Dirt' in this case does not only mean solid dirt particles but also other contaminations such as oxydation products, water etc.

bar

The extent to wh~ch a transducer's sensitivity to dirt is important, depends on the systems in which the transducer is applied: 'research systems'more often then nothave good hydraolie liquid filtratien meaning that sensitivity to dirt will be of minor importance.

In most 'production tests~ a much smaller sensitivity to dirt ~s imposed because of the inferior filtratien eapaeities of these systems (absolute degree of filtration: 40- 60 ~m).

In 'field measurements' very often an almost absolute insensitiv-ity to dirt is required whieh eonsiderahly limits the number of suitable transducers.

From the floor, the following remarks eoncerning sensitivity to dirt were made:

According to speaker's apinion the problem of sensitivity to dirt will solve itself: since pressores applied in hydraolie systems are constantly increasing, the filtratien capacity has to increase as well, not because of the transducers' sensitivity todirtbut ~n order to garantee correct operatien of the hydraolie eomponents. This remark was accepted as correct as long as the transdoeers do

not require a better filtration than the hydraolie components do, ~n other words, a filtratien not exceeding anproximately 10 ~m.

20

-A number of transducers, however, requ1re a much higher degree of filtration, for example the hot \.ire anemometer: with an

absolute degree of filtration of 0,5 ~m(!). Seen from this angle, the remark was considered as too optimistic, also for the (near) future.

f. Energy dissipation: The only remark made about energy dissipation went to say that the smaller the energy dissipation, the better

this 1s for the choice of transducer.

g. Purebase costs: Reference was madetoa list of approximate purebase costs for transducers and processing instrumentation for a number of measuring principles as given on page 40 (varia-tions of 10- 20% in tbe given purebase casts are possible).

h. Stability: The following definition was given:

Stability

=

Tbe time allowed between two subsequent calibrations in whicb the accuracy of the measurements (based on tbe calibration) can be kept witbin a given value.

1, Life expectancy: Life expectancy was cbaracterised by tbe Cbairman as being a very important point, regarding wbicb little is known, neitber from tbe manufacturers 'side nor from the purcbasers' (e.g.

his expectations).

As appeared from the discussion, vibrations and shocks in the sysbem are some of the most important factors influencing life span. Here, a request for a shock resistance of 50 g was put forward for cases where no special provisions have been made

(i.e. no special suspension, damping etc.).

J· Dynamic properties: No further remarks were made as toa

sub-division of transducers into first and second order systems, these

being respectively characterised by time constant, natural frequency and damping factor.

From the floor came some more proposals for additional selection parameter for volume flow and mass flm.;r transducers:

k. Introducibility: How easily can a transducer be mounted 1n a exis-ting system without pre-empexis-ting the system?

l. Shock resistance: To which extent can a transducer resis-t system induced vibrations and shocks (see a lso point i) life expectancy).

m. Rigidity: How rigid is a transducer?

n. Sensitivity to overlaad and shockload: This parameter was added because in case of trouble shooting (field measurements), one very aften does not know neither dimensions nor type of flow

to be measured.

o. Sensitivity to the arising of static electricity: Can static electricity occur and if so, what is its influence on measure-ments?

p. Sensitivity to pressure and temperature Sh!~g~! durin& the measurements

q. Processibility: How easily can the output signal of the transducer be processed?

r. Sensitivity to turbulence: Do turbulence and swirl influence measurements?

s. Is reverse flow possible: Can the transducer be damaged by reverse flow?

22

-Theme 4

What type of measurement devices are suitable for these fields of application in view of their properties as compared to the required properties agreed upon in item 3?

Opinion of the Power Transmission Department

Without wishing to cover the entire scale of possibilities. we'd like to draw your attention to a number of the best known measurement-principles including some new developments in this field as well.

We would like to discuss the following flow meters: I • the turbine meter

2. the positive displacement meter 3. the float meter

4.

the restrietion meter 5. the pit ot-tube

6. the swirl meter

7. the drag force meter

8. the hot wire- and hot film anemometer 9. the electro magnetic meter

IO.the ultrasonic meter ll.the Laser-Doppler meter 12.the 'fluidic spring' meter 13.the boundary layer meter

Re I. The turbine meter. The functioning of this meter is based upon measuring the number of revolutions of a flow-driven turbine \-Theel; whereby the number of revolutions acts as a flow indicator.

This meter may have good qualities over a large measuring range (1:10) as regards accuracy and repeatibility depending upon whether or not this meter is to he used as a linear one (i.e. whether or not a linear relation between output signal and flow

is required).

The range in which linearity exists (within a predetermined accuracy) decreases as fluid viscosity increases. For instance: for a dynamic viscosity of 60 cP, the linear range will only be 1:2,within a required accuracy of~ 0.5%.

By using the meter as a non-linear one as well - which hence implies calibration of the meter - a relatively large range is obtained with the same accuracy of + 0.5%, and a repeatibility of 0.1%. One should notice that a larger linear range or a larger viscosity range may be obtained whenever less accuracy is required. It also holds true that it is sensible to measure well over and above the lower limits of the range, especially when high accuracy is required, to avoid low repeatibility. The fact that repeati-bility decreases rapidly when reaching the lower limits of the working area, is probably due to the flow at the turbine wheel being in unstable equilibrium between laminar and turbulent; an equilibrium which to a large extent depends on accidential in-fluences.

The influence of viscosity upon the behaviour of the turbine meter is considerable. Not only do the calibration curves change with viscosity (think of the decreasing linear range), the total range

24

-with increasing viscosity. Hence, the maximum admissible pressure drop, which depends upon the maximu~ acceptable load upon the hearings, wi ll be reached sooner.

Viscosity dependenee of the meter is considerable, counteracted by the fact that measurement-results do not depend upon fluid

density, as this is a "olume flow meter. Maximum admissible pressure and temperature will present no problems meaning that this turbine meter may be used both in high and in low pressure systems, in-cluding return- and suction lines. When, however, applying a meter in return or suction lines, one has to uroceed carefully as

pressure-drop over the meter is considerable and - at maximum

flow- may reach 3.5 to 4 bar which is of the same order as nressure in return- and suction lines. Such a pressure-drou has the further disadvantage of leading to considerable energy dissipation.

The following may be said as to sensitivity to dirt:

the turbine ·meter ~s sensitive to dirt in so far as turbine meters ~ith electro-magnetic speed sensors are concerned: here the uresence of a magnet - either in the turbine wheel itself or in the nick-up spool - attracts ferrotic dirt particles, leading to a pile uu near the turbine wheel, quite apart from the fact that the 'fibre-like' dirt,present in the shape of longstrandsin the hydraulic liquid, will be caught by the turbine wheel. The pile uu of dirt particles may influence the meter's life expectancy which is determined by the hearings of the turbine wheel: these varticles rnay cause abrasive wear. In general, the life expectancy of these hearings and hence of the meter is good, that is if the maximum hearing load is not exceeded, although even short overloading of

150 to 200% does not do any harm.

The rneter's long term repeatibility is exceutionally high, which is corraborated by the fact that the calibration deviation for several turbine meters has not changed by more than 0.1% in most measuring points over a year of - non continuous - use.

As a last item we would like to discuss the dynarnic ~roperties of the turbine meter:

Dynamically seen, the turbine meter is a first order system in which the time-constant is mostly determined by the moment of

inertia of the turbine wheel, and ly the magnitude of the flow as well. 'Small' turbine meters intheupper part of their range have a time constant such that - to a limited extent - they will be approprlate for measuring varying volume flows.

Re 2. The positive displacement meter.

fig~

The functioning of this meter is based on the principle that per rotation a known, unchanging volume passes through the meter (where-by leakage 1s neglected), so that the number of revolutions of the meter's shaft is an (almost) linear measure for the flow. This meter has very good accuracy and repeatibility orovided .at the sneed of the shaft can be measured accurately. For an ov _ wheel meter, for instance, the following values apply: accuracy of + 0.5%, repeatibility of 0.1%, within a range of I : 10.

In a more limited range even bettervalues may be obtained: up to ~ 0.1% for accuracy. If one uses a niston tyne motor as a

positive displacement meter one has to proceed carefully: when used in high-pressure lines of a system, the leakage clearances

26

-will show a pressure difference equal to the pressure 1n the system itself. This means a relativel• large leakage: hence the difference between real and measured value is considerable as well. As far as maximum admissible pressure is concerned, these meters can only be used in a limited way. This problem does not exist for other meters, like e.g. the oval wheel meter: here, the pressure drop over the clearances is equal to the pressure dron over the meter itself. The only thing that could happen tothese meters 1s

that they may expand a bit. The maximum accuring operating

temperatures however,never prevent the use of positive displacement meters in hydraulic systems.

The influence of density and viscositv,on the behaviour of this type of meters,which lead to a change in leakage flow, may be called small, as is shown by the high degree of accuracy. The positive displacement meter has a low sensitivity to dirt. As the construction of this meter corresponds to that of many other

components in hydraulic systems, special protective filtering will not be necessary.

As to energy dissipation 1n this type of meters, we have to dis-tinguish again between specially constructed positive displacement meters and the piston motors used as such. In the first case, pressure drop over the meter is rather small and hence the energy dissipation as well. In the second one, however, both energy dissipation and pressure drop may become considerable; this, 1n

turn, depends on construction (internal resistance).

Apart from relatively low purebase costs (excluding speed sensor), the positive displacement meter has a practically unlimited long-term repeatibility and long life expectancy, thanks to its rigged construction. One important feature of the positive disvlacement meter is its degree of _non-regularity resulting

in a non-regular angular velocity of the output shaft, whereby considerable deviations may occur between real and measured values during instantaneous measurements.

As to dynamic behaviour, we can be brief:

The positive displacement meter is not suitable for measuring

1. The moment of inertia is very large, resulting in a very large time constant of this first order systern.

2. The non-regu lar i ty of the meter, c:ue to which f act the angu lar velocity will show neriodical changes even at constant volurne flow. These changes denend on the nurnber of revolutions.

Re 3. The float meter

::

:. :-.·.·· ... . :. : . :

:

. · .. :

·

.

.,.

.

·

.. .

···

.. x x(Q ) V

The functioning of this meter 1s based upon rneasuring the

position of a floating body in a tube with variable cross area. The position of this floating body is in rnany cases an alrnost linear flow rneasure. For this type of meters a relatively close correlation exists between range and obtainable accuracy. By changing the

apparent weight (that is real weight minus lifting force) the range can be varied. However, by enlarging the range, both sensitivity and accuracy decrease, so that an optimum between range and accuracy has to be sought. In practice, one rnostly selects a range of I : 10, with an accuracy of~ 1 to + 2% and a repeatibility of~ 0.5%. Another important item is, that the meter has to be rnounted perpendicularly for optirnal and accurate

28

-is both density- and v-iscosity depender:_, although the latter can he decreased by using a specially sr ;Jed floating body. The

dependenee on density is unavoidahle, since·it directly influences the apparent weight of the floating body. The most common design of the float meter has a conical glass tube, meaning that temoer-ature and pressure may create problems when applying this float-meter to hydraulic systems. These problems are solved by means

a specially designed metal tube, whereby the position of the floating body is determinated magnetically. The disadvantage of these special designs versus the more common glass tube lies in their high price, the more so since precisely the glass tube float meters are cheap in comparison to .others.

of

An advantage of the float meter resides in the small and flow-independent pressure drop over the measuting device, so that only a small energy dissipation will occur. Further there is only a small sensitivity to dirt, a long term repeatibility which is nearly unlimited and a long life expectancy (no rnaving parts establishing any contact). As to dynamic qualities, the following may be mentioned: dynamically seen, the float meter is a second order system with a very low natural frequency and a low damping factor, which make it

unsuitable for measuring varying flows.

Re 4. The restrietion meter. The functioning of this meter is based on measuring pressure difference over a restrietion in the volume flow, whereby this pressure dron is a measure for the volume flow through this restriction.

Discussion will be limited to three types of restrietion meters with fixed cross area; that is the orifice, the measuring bow and the laminar flow meter, the latter being an exception to the other meters.

A. The orifice

pressure transducer fig~

The orifice is for many reasons, the most widely used flow

meter: exceptionally low purchase cost (excluding the differential pressure transducer); the many details known about orifices; its simplicity; the fact that calibration becomes superfluous when standardized orifices are mounted; its nearly unlimited life expectancy; lew sensitivity to dirt; and good long-term

repeatibility. The only influence dirt particles may excercise on life expectancy, long-term repeatibility and accuracy is that they erode the sharp edge of the restrietion opening, so that accuracydecreases and calibration curves start wandering, which

on the long run will limit life expectancy. Compared to other flow meters, however, this influence is minimal.

The accuracy and repeatibility obtainable with the orifice (which of course, also depend on the accuracy of the differential pressure transducer) depend on the accuracy with which fluid viscosity and density may be measured and kept constant.

Viscosity and density determine,by means of the Reynolds-number, the flow factor a an experimental correction factor for the

relation between pressure difference and flow for a given orifice. Pressure and temperatures ~n hydraulic systems do nat exclude the application of this meter ~n these systems. Disadvantages of orifices consjsts of the relatively small measuring range of I : 3 up to maximum I : S,the high permanent .pressure drop over the meter which - depending.on the opening ratio - may amount to 50 - 95% of measured pressure difference which results in

30

-high energy dissipation. The dynamic rroperties - which depend on the dynamic qualities of the added pressure difference transducer-are such that the meter may be used for a reasonable frequency range; in which case an orifice with pressure transducer openings close to each ether is necessary in order to restriet phase

differences between the two signals (orifice with 'corner taps for pressure measurement').

B. The measuring bow

öl) ~ 0 .-I td Q) !-I ;:::l U) U) Q) ~

Pz

I • exter1or wall

permanent pressure loss over tbe elbow

wall

In this case,pressure difference is measured between the inner and outer wall of an elbow. Calibration is necessary in order to obtain rather àccurate measurements, though accuracy will never be as good as that of the orifice, sirree the flow pattern is too unsteady which also explains the none too good repeatibility. With good calibration a maximum accuracy of .!_ 2% is obtainable combined with a repeatibility of maximum.!_ 0.5%.

A disadvantage lies in the small pressure difference over the meter, which makes a sensitive and therefore expensive pressure difference

transducer necessary. An advantage of this small pressure. di fferencp ltowever, will he a small nermanent pressure loss as compared to the orifice. Both meters however, have the disadvantage that· for correct operation long sunply- and discharge lines will he

necessary in order to obtain an undisturbed flow pattern. We cannot say anything as to the dynamic properties of the measuring bow as we are lacking experience in this matter.

C. The laminar flow meter

• I

stra~ghteners

/metering element

fig~

The functioning of this meter .is based on the law of Hagen-Poiseuille, which indicates ,that the re lation between flow and pressure drop over a tube is linear in case of a viscous, imcompressible fluid.

It should be stated that flow through the narrow capillaries

~n case of hydraulic transmissions will not be isoviseaus so that the linear relationship will be lost.

Apart from this, the meter oossesses a number of other disadvantages such as: very high purchase casts; a very high sensitivity to dirt

(the narrow capillaries easily get clogged)i a very high pressure drop ( the total measured oressure difference, which would be ver;v high because of high viscosity of hydraulic liquids, would be

lost); and very large dimensions. Because of these disadvantages, we think that this meter is not sui table for application in

32

-Re 5. The pitót-like tube. This includes bath a pitot tube and

'an Annubar 'low meter.

A. The pitot tube.

static taps

I

Z::z::l '--·

~o

differential oressure transducer

We think that the pitot tube is nat suitable for application ~n

hydraulic systems for the following two reasons:

I. Due to the high viscosity of hydraulic fluids additional viscosity farces will occur at stagnation point leading to major deviations in the measured pressure difference ~s one measures the difference between the total pressure and the static nressure).

2. The diameter of the nitot tube is such that when using a nitot tube in the lines hydraulic systems, the cross. at;ea will

considerably decrease whereasvelocity will increase very much. which will result in a completely wrong measurement.

B. The Annubar flow meter I up-ele s ream/ ment

c

p ~interpolating ~ tube '--down-s tream element

J

We have not had any experience with this meter. We have, however, had rather enthousiastic reactions about this meter from the United States. We would like to hear contributions from those of you who have any experience with this meter.

Re 6. The swirl meter

ortex precession rl component

We have had no experience here either and wouldlike to call upon those of you who have.

34

-The following may be said as to dynamic properties of the swirl meter: the manufacturer of this met' ~ gives a time constant of one second for gas application. The reason for this probably lies in the very complicated flow pattern which will need a long time to reach a new stationary equilibrium 1n case of flow change. Hhen fluids are measured, we may exnect to see an even higher time constant (due to larger density and inertia). For this reason, the meter will he completely unsuitable for measuring changing volume flows.

Re 7. The drag force meter

\

~,...

"

_"'

1\ r-strain g a ge r-hollow t elastic u he beam bl_drag body ) fig_. _1_1

The functioning of this meter is based on measuring the drag force (by means of strain-gauges), exercised on a drag force body by the flow. The obtainable accuracy of this meter is + 0.5% or better, at a

repeatibility of~ 0.1% 1n a range of I: 5 (obtainable due to the approximate square-root relation between flow and output signals),

provided the meter is calibrated. The reasou for this calibration

lies in the fact that the drag force t;oefficient is not constant. but a Eunction oE tlle l{eynolds nurnber, and thus of the fluid 's velocity, vis-cosity and density.Furtherrnore,density will influence drag pressure

0

p v2, where p

=

density and v

=

veloei ty), hence, this meter has a relatively large density dependence. The temperatures and pressures accuring in oil hydraulic systems do not exclude application of this meter in these systems.

Advantages affered by this meter are: its very large insensitivity to dirt; its good long-term repeatibility (due to the use of a strain gauge-bridge for measuring drag force); it's long life

expectancy (due tonon-contact of rnaving parts); and the relatively small energy dissipation (due to a small pressure drop). Purchase casts of this meter are rather high, which is a disadvantage.

As to dynamic properties of the drag force meter, we can state that this is asecondorder systern with a natural frequency of 150 to 200Hz (however, drag force meters exist with a rnuch higher natural

frequency). Hence this meter may be used for rneasuring changing flow in a limited frequency range. However, one has to see to it that this meter is not tuned to its natural frequency range (which also goes for stationary rneasurements, what with the pumP rib).

Re 8. The hot wire and hot film anemometer

D.C.-amplifier

36

-The functioning of this meter 1s hased on discharge through the fluid to he measured, of heat developed in a thin wire or film, this discharge being dependent on and thus constituting a measure for the velocity of the fluid.

In fact, this meter has been especially developed for measuring dynamic flow phenomena, where accuracy is of secundary importance. This meter is hardly suitable for measuring stationary flow.

Viscosity- and especially density dependenee are enormous, as well as its sensitivity to dirt. Long-term repeatibility is very poor, due to the very large sensitivity of the heat-exchange coefficient

to external circumstances, while life expectancy of the trarisducer will he very limited (hence this meter surely can nat he used for

continuous application). Temperatures and pressures, which

normally exist in oilhydraulic systems will cause a lot of problems in commercially available roodels of this meter though by now special designs for hydraulics do exist.

Because of its lesser vulnerability, a hot film is preferred to a hot wire as transducer element. The advantages of this meter are: its very small energy dissipation; its very large range (un to

I : 100; 1n which case the sensitivity of the meter will he very small); and especially the excellent dynamic properties which

allow for measurements up to very high frequencies.

An

important aspect however is the fact, that this meter really is a velocity-meter

only measuring changes in local velocity unless special provisions are made for local velocity being representative for the total flow.

Re 9. The electro-magnetic flow meter

The functioning of thi~ meter is based on the following principle; over a conductive liquid, which flryJs through a homogeneaus magnetic field, a voltage drop will occur which will be proportional to

the fluid's velocity if the electric conductivity of the fluid is high enough.

This meter, which possesses a number of good properties (among others very good dynamic ones) is, up to now, not suitable for

oilhydraulic systems due to the fact that the electric conductivity of the hydraulic liquids is too low.

Up to now, efforts to adapt this meter to oilhydraulic systems (by increasing the conductivity by means of dopes) have failed, because these dopes proved to be very aggressive to other comDo-nents of the systems (e.g. seals).

Re IO. The ultrasonic flow meter

~S:::S:::S~U:::S:::~/, .6;::s:~~ / / SI S2: V = c + v

}

I c 1 _(c + vc)ti (c - vc)t₂ s3 - S4: v₂= c

- v

_c v 1 _{(-I- _I_)}

v

c

fi~ cosp COSj3 tl t2

The functioning of this meter 1s based on measuring propagation time of an ul trasonic signal through flow. Th is meter is, un to now not suitable for oilhydraulic systems because dimensions of the transmittors and receivers are such that the smallest tube diameter in which they can be used is 300 mm, which is much larger than

38

-the usual oil hydraulic tube diameters. Fur-thermoré, purchase costs are so high that one would 1 Je doubts as to their anplic-ation possihili.ties ~n oil hydraulics, the more so since

obtainable accuracy ~s not very excellent.

Re 11. The Laser-Doppler anemometer

unit ohoto multiplier

The functioning of this meter is bas~d on measuring the Dopnler frequency of the Laser light, which is strayed at the ooint of measurement by the solid particles present in the fluid - this ~s clone with a photo-multiplier. The Power Transmission Department does nothave any experience in the use of this meter, which is one of the latest developments in flow measurement. Therefore we would like to call upon those amongst you who have

experience in the use of this meter.

Re 12. The fluidic spring meter

The functioning of this meter is based on the oscillation of a mass by flow, the frequency of this 0scillation being pronortional

to the volume flow. This meter is & recent development and not yet in production.

However, prototype tests of this meter show this to be a very interesting development.

It remains to be seen to what extent this meter will be suitable in oilhydraulic systems.

Re 13. The boundary-layer meter

temperature profile befare the heater coil

fig~

boundary layers

temperature nröfile behind the heater coil

The functioning of this meter is based on measuring temperature drop over the boundary layer of a flowing liquid. This dron is brought about by heating the fluid through a coil placed around the tube. The Power Transmission Department has no experience in the use of this meter. We e~pect that this meter is rather similar to the notwire and hotfilm anemometer. In view of the very limited thickness of the boundary-layer (i.e. a small thermal capacity), dynamic behaviour will be excellent, though, as for the hotfilm meter the question arises to what extent the boundary-layer is representative of the entire volume flow.

40

-Approximate purchase costs for transducers and instrumentation*

Measurement principle I transducer instrumentation total

hot :film anemometer 1000 3500 - 4500 4500

-

5500

orifice :from 100 2000

-

3000 2100

-

3100

f'loat meter 500.- 600

-

500

-

6oo

drag :force meter 3000 1500 4500

oval wheel meter 1000 mech. 200 ~* 1200

electr. 1500 2500

hydraulic motor 1500

-

2000 mech. 200 1700

-

?200

electr. 1500 3000

-

3500

turbinemeter from mech. 200 800

_-

900

600

-

700 electr. 1500 2100

-

.'22 00 swirl meter 7000 3500 10500 electro-magnetic 4000

-

5000 4ooo

-

5000 meter

I

ultrasonic meter 20000 20000 (s.A.-principle)

I

Laser-Doppler meter 55000 55000 ..,.. Prices in D.l\1.

**

mech.

=

mechahical instrumentation electr.

=

1~lectronic instrumentation

Summary of discussions:

For reasons of time, it was decided to have a short

intro-duetion only,limited to those measurement principlesas have been previously described ~n the above mentioned 'Opinion of the

Power Transmission Department'. From these in turn,· only those principles which are applicable to oilhydraulic systems. have been selected. i.e.

A. Steady state measurements: turbine meter

positive displacement meter drag force meter

float meter

restrietion meter

B. Dynamic measurements: restrietion meter drag force meter hot film anemometer Laser-Doppler anemometer turbine meter

42

-Theme 5

Which flow transducers are conside:cd to bemost suitable for measuring steady state flow (mass flow) in oilhydraulic systems. in view of the requirements as discussed sub item 4 ?

Opinion of the Power Transmission .Department

In view of the properties discussed sub item 4 , the following meters are most suitable for measuring steady state flows in hydraulics.

1. The turbine meter: lts properties make this meter suitable for more than one field of application; attainable accuracy and repeatibility in calibration are such that this meter is suitable for research purposes, as well as for production tests.

This meter may be used for field measurements because of its fairly large measuring range (what with the required accuracy), its very good stability, long life expectancy, and its none too excessive sensitivity to dirt. The relatively very high pressure-drop,however, is a disadvantage.

2. The positive displacement meter: This meter will, because of its properties, be suitable for production tests and especially for field measurements in view of its large measuring range, stability, life expectancy and relatively low purchase costs. Disadvantages are pressure drop over the meter which may be considerable, plus the fact that this meter is not suitable for high pressure ranges.

3. The float meter: This meter will be suitable for field measurements because of its large sensitivity to dirt, stability, life exvectancy, small energy dissipation and

relatively large measuring range. A disadvantage is that 1n the standard version.pressures and temperatures such as they occur in hydraulic systems may create problems, quite apart from the fact that it will always have to be placed in a vertical

4. The drag force meter: Like a turbine meter. This meter is suitab~

for more than one field of application. Sufficient accuracy and repeatibility may be attained ·for research purposes and this meter is also very well suitable for production tests. This meter is suitable for field measurements because of its great

in-sensitivity to dirt, r,tability, long life expectancy and low pressure drop over the meter. Disadvantages are: the high

purchase costs and the small measuring range. Further, for steady state measurements, it should be born in mind that this meter should not be tuned to its natural frequencies.

5. The orifice and measur~ng bow: These meters are likewise suitable for hydraulic systems, although less so than the above-mentioned volume flow meters: this is due to their very large viscosity-and density dependenee plus the fact that very long supply- viscosity-and discharge lines are necessary.

44

-Summary of discussions:

In order to discuss the transducers selected sub item 4, we used a table, in which qualification characteristics are given for each and every transducer, based on the selection

parameters agreed upon sub item 3.

For each and every transducer, this table is g~ven foliowed by discussion.

Furthermore, it has to be noted that no further explanations were given as to the transducer's construction, an exception to be made for the hot film and the Laser-Doppler anemometer, since these transducers represent two new, little known principles. As these transducers are especially suitable for dynamic measurements, a description will be given sub item 6.

A. Turbine meter

accuracy: ~ 0.5% when calibration ~s possible

viscosity·and density dependence: viscosity dependent (disadvantage) range: 1:10

maximum permissible pressure and temperature: no limitations sensitivity to dirt: sensitive to ferrotic particles

sensitive to fibre-like dirt

energy dissipation: rather high (up to 3,5 - 4 bar at max.flow) purebase costs: reasonable

stability: good

life expectancy: good

dynamic properties: moderate (perhaps too optimistic?) shock resistance: reasonable

introducibility: bad rigidity: good

overloading: 150 - 200% in differential pressure is possible sensitivity to the aceurenee of static electricity: no

sensitivity to pressure and temperature changes

processibility: very good (frequency-analogue output) sensitivity to swirl: moderate

reverse flow possible: yes (some constructions are bi-directional) comT?atibility to synthetic fluids: good

Connnents

The following remarks were made regarding this transducer: The Sensitivity to ferrotic particles can he avoided by using an electrooie speedometer, in which case no magnet is present which might act as a magnetic filter.

B. P.D.-meter

accuracy: + 0.1% possible (as result of the displacement

principle)

viscosity-and density dependence: rather small range: I: 10 with an accuracy of.:!:_ 0.5%

maximum permissible pressure and temperature: low to very low pressure

sensitivity to dirt: reasonable energy dissipation: moderate purchase casts: reasonable stability: very good

life expectancy: good

dynamic properties: very poor (through inertia of rotating elements) shock resistance: resonable

introducibility: very poor rigidity: good

overloading: 10-20% (of max. allowable speed, thus flow)

sensitivity to the accurance of static electricity: no sensitivity to pressure-and temperature

changes during measuring: yes

processibility: with electrooical speed transducer very good

46

-sensitivity to swirl: no reverse flow possible: yes

compatibility to synthetic fluids: moderate

Discussions

The ensueing discussion clearly showed that the different P.D.-meters' properties depend mainly on the 'construction principle' on which the meters are based.

The above-mentioned data reflect the properties of the often-used aval wheel meter. Other P.D.-meters, however, can show

totally different properties: hydromotors, for example which are aften used as P.D.-meters. These hydramotors have a small leak flow over a rather large pressure range, and hence are suitable for bath high pressures and a very large range; for example, a radial ball piston motor which can be used up to a

system pressure of 300 bar and over a range of 1:100, with an

accuracy of ~ 0.5%.

A disadvantage, however, ~s that pressure drop over the meter increases with increasing range whilst high pressure drops are aften nat admitted, nat even ~n measuring systems.

However, in many cases where pressure drop over the transducer is of minor importance, hydramotors are applied for flow

measurement, although there again not all types of hydramotors are suitable for flow measurements. Mostly, piston type motors (axial and radial) are used for low pressure differential conditions, where 'low' sametimes can mean only 7 bar, it being understood that accurate calibration of the hydromotors, when used as flow transducer is recommended.

It was further noted that for several P.D.-meters temperature limits apply (to avoid mechanical damage to the transducer).

C. Float meter

accuracy: ~ I - + 2%

viscosity-and density dependence: particularly sensitive to density range: I: 10

max. permissible' pressure and temperature: for normal construction:low sensitivity to dirt: low

energy dissipation: low purchase casts: low stability: very good

life expectancy: very good dynamic properties: poor shock resistance: poor introducibility: very poor rigidity: moderate

overloading: nearly unlimited (in flow)

sensitivityto the accurance of static electricity: yes sensitivity to pressure.and temperature

changes during measuring: no

processibility: for normal construction: very poor

for special construction (magnetic piek- up with electric voltage output): moderate

sensitivity to swirl: no reverse flow possible: no

compatibility to synthetic fluids: good

Discussions

During the discussions, the following three points were brought up: I. introducibility

2. sensitivity to dirt 3. processibility