Note on a datalogging system with engineering research : optimizing of research efforts

Note on a datalogging system with engineering research :

optimizing of research efforts

Citation for published version (APA):

Veenstra, P. C., Heuvelman, C. J., & Hulst, A. P. A. J. (1967). Note on a datalogging system with engineering research : optimizing of research efforts. (TH Eindhoven. Afd. Werktuigbouwkunde, Laboratorium voor mechanische technologie en werkplaatstechniek : WT rapporten; Vol. WT0183). Technische Hogeschool Eindhoven.

Document status and date: Published: 01/01/1967 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.

titel:

auteur(s):

sectieleider:

hoogleraar:

Note on a Datalogging System with Engineering Research. Optimizing of Research Efforts

P.C. Veenstra C.J. Heuvelman A.P. Hulst

sa menvatting

prognos.

A description is given of an instrument for automatic recording of process data, which is in use with the Laboratory for Production Engineering of the Technische Hogeschool at Eindhoven. Besides attention will be paid to

the analysis and elaboration of the data recorded.

tr.fwoord: datum: July 12th 1967. aantal biz. gesc:hikt voor

'-publicatie in: ~ote presented o.C.LR.P.

1.

2

-Note on a Datalogger system in.use with Engineering Research Optimizing of Research Efforts

Introduction

P.C. Veenstra C.J. Heu~elman A.P. Hulst A data logger is an instrument for automatic recording of process data.

In this note the data logging system which is in use with the laboratory of PreductionEngineering of the Technische Hogeschool at Eindhoven will be described. Besides, attention will be paid to the analysis and elaboration of the data retorded.

The application of' automAtic data logging proves to have a.number of advan-tages:

1. The recordings are 'made by the system during the experiment; the operator can pay his full attention to the process itself.

He loses no time for reading the measuring data.

2. Owing to the relatively high speed (in this case three readings per second) more data can be recorded; this isf,g~importa:p.ce both in semi-dynamic

processes and in statistic evaluation of process data.

3. The chance of human shortcomings like accidental or systematic personal errors and mistakes when reading the data is eliminated.

4. The recorded data, which are punched in a tape,can be directly presented to a digital computer.

The time wanted for the tabulating and regrouping of data with a chance of making mistakes is· greatly reduced; however, 'a suitable computer programme , is needed.

The· disadvantage of automatic data recording lies in the fact that human supervision of the data measured is difficult, so that it is possible that the datalogger records misreadings without warning.

The datalogger in use is

a

rather universal instrument, and is not committed to one type of process. It is clear that the most important application is in processes where a great number of variables is to be recorded repeatedly. 2. General Description (figure 1)

In every test car:a;ied"out to study the course of techno~qgical processes a distihttion can be made :petween a number of indep~nd~nt,. process variables set at chosen values and .. fFuumber of dependent variables determined by process conditions. '

Thus, in metal cutting (turning) the cutting speed, feed rate, chip geometry, and a'humber of material constants are the independent variables.

Dependent variables are cutting forces, temperatures, rate of deformatiP. and rate of tool wear, etc.

In connection with the data logging equipment a different classification can be made: .

A. Variables of w~ich,th~ magnitude is directly available either as an analogue voltage (e.g. cutting forces: strain gauge bridges; temperatures: thermo-E.M.F.'s) or as a digital signal (e.g. spindle speed, feed rate: frequency counters).

B. Variables of which the numerical value can be determined only at the end of the experiment (deformation, wear) or variables of which no suitable transducers are available to transform the value into an electric signal (angles, dimensions).

During the measurement the magnitudes of variables of type A are stored successively in a·number of scans punched in paper tape, those of type B are punched by hand before or after the experiment. In figure 1 a survey is given of the complete data logging system. The electronic part of the system consists of: an automatic switch

which can scan,a number (maximum 20) of different channels successively, a digital voltmeter which converts the analogue voltages into digital information, and ~,a, puncher driven by an output control.

Information of type B,is punched by the operator with the aid of a device on the output control (second source input), The operator can select the nUIl'l.ber of channels of variables of type A. After a

starting signal the instrument scans,the selected channels continuously and the measuring results are punched. At the end of an experiment a stopping signal is given.

The punched tape nowcoutains one series of data of type B and a number of series qf type A. The tape is fed into a digital computer which, with the aid of '¢alibration tables, transforms the measured

signals into figures with the correct unity.

Besicies those "primary" variables it is often wanted to know a number of derived "secondary" figures which are of importance in the process, The computations invo.l.vE;d are carried out by the computer which delivers a secon~ pauchedtape (tape No.2) which contains the primary and

secondaryvariabl~s in the desired form.

The routine of the mathematical analysis of the data obtained depends on the ailll!'of the' experi~l(!nt.

In the case where a theoretical model of the process is available, theoretical predictions can be compared with experimental data by using a suitable computer programme.

If a reliable model is lacking, the computer can produce empirical or semi-empirical relationships as a stepping stone towards theoretical analysis. As to this, the use of a plotter which automatically draws the graph of any variable in relation to another is very helpful. 3. The Electronic Equipment (figure 2)

In section 2 it has been mentioned that two kinds of variables exist, type A and type B. The first are represented by analogue voltages and must be converted into digital form, while the others are already

presented in digital form with the aid of digitally adjustable switches (second source).

A mote detailed d~scription is given below. 3.1 The Input Panel

A number of' at mos,t 20 different signals in analogue form can be connected to the equipment at the input panel. Below each input connector a digital switch is mounted, by means of which the digital voltmeter can ,be programmed at one of the eight available different ranges. This makes it possible to select any individual range during the same scan in the various channels, so that voltages down to 1 pV and up to 750V can be recorded with the highest possible accuracy. This makes the equipment very flexible in use.

4

-3.2 The Scanner (Honeywell type 707A-l20tt])

In actual fact the scanner is a stepping multi-switch, which connects successivttly any of the 20 input signals to the digital voltmeter. In each channel six wires hayeto be switched: three for the signal and three for the programming command. At the scanner panel the lower and upper signal limits of the channels to be scanned can be selected by meapa of a digital .switch. After a starting command (given by hand or remotely by an external source) the instrument begins to step auto-matically, starting from the lower limit. The scanner can make either one single scan or a continuous series of scans depending on the choice made by the operator. The stepping action can also be controlled manually.

On. a visible readout the operator can see which channel is being switched. The multi-switch needs at mO$t 8ms for switching from one channel to the next.

3.3 The Digital Voltmeter (Honeywell type 630 S (2)

This instrument converts the. analogue d.c. voltageVin. into a digital signal. The voltmeter is of the integrating type and hence the output signal Eout is proportional to:

f

E out=-

~

f

V. dt

1n '

o

where T is the integrating time, also called the encoding time. This time can be set either,at O.ls or at Is. The advantage of the integrating voltmeter over other types is that superimposed noise and interference

signals are rejected to a great extent. The accuracy of the voltmeter is IOOppm of the encoded value plus 10 to IOOppm of the selected range, depen-ding on that range. The output is in binary-decimal coded parallel form and is stored in a memory after each encoding until the next encoding is finished. The encoding command is given by the scanner just after a switch closure. 3.4 The Output ContTol(Honeywell 825-C2 (3J)

The function of the ou~put control is to drive the puncher after transferring the digital parallel information into the series form. So each character is fed successively into the puncher. In this case the IBM-flexowriter code is used.

In order to represent a channel a so-called "word" is punched, each word consisting of seven successively punched characters:

character 1 and 2 channel number character 3,4,5 and 6 magnitude

character 7 end of word sign

For expressing the magnitude of the signal 4 decimals are in use, so an accuracy of at ~bes.t:: ) part in 10,000 can be obtained. However, the code format of the digital voltmeter is longer (6 decimals) but in the majority of experiments such a high accuracy is not required. It is possible to change the code format of the output control. If the system makes an error

(e.g. overloading) the least significant decimal is replaced by an error symbol (character

6).

The computer then skips the whole scan in which the error occurs.

The output control staxts cenverting after the "encode complete" command from the voltmeter. (1s the output control drives the puncher this sigJJal is also'called tpe "punch cemmand";,

3.5 The Puncher (Friden model.,S.P.-2)

4.

4. 1

The puncher is drivenqy the output control and punches 2S characters per second, so with a word format of 7 characters it takes 280 ms for punching a word.

When a word is punched the output control gives a "punch compl.ete" signal to the scanner which· then steps to the· next channel and so on. With a stepping time of 8 ms .and an encoding time of 100 ms, for each channel a measuring time of 338 ms is required.

Elaboration of the Test Results

As stated in chapter 2 two cases can be distinguished:

1. Testing of theoreti~al relations between the various process variables. 2. Tracing a relation between the variables without a theoretical model. Testing of a Theoretical Model

In this case which. wiJ::h respect to the elaboration, is more simple, the recorded values of the most important variables are to be compared with the values calculated from a theoretical model with the aid of a digital computer. With simple variance analysis, possible significant differences of the values can be traced. Sometimes the constants of an existing model are unknown or not s'l,lfficiently accurate. In these cases it is possible, with the aid of a regression progra.nnfte, to derive the constants from the test results and at the same to get an idea of the reliability of the model.

4.2 Tracing an Empiric Model

Very often it is desired to derive an empirical mathematical relation from the test results; this relation then gives the results of the investigation in a practical and closed form. However, owing to the absence of insight in the physical background of the process it i~ difficult .to ... find the correct form of such a mathematical relation.

In general it is not possibl~ to give rules for the choice of a mathematical expression. In order to give a theoretical interpretation it will be prefer-able to find a simple relation; such as line~ square or hyperbolic functions e-powers, logarithmic or exponential relations between two variables. The double logarithmic functions in the form

log:u= a log X + b log y + clog z + d

or

a b c

u

=

x.y Z.K

are handy for empiric use; the disadvantage of this relation is.that later on a physical interpretation of the process becomes very.difficult. A clear survey of the possibilities for the research-worker is given by Mandel

[4].

I f a relation is chosen with the aid of a suitable regression pregramme[S] the unknown constants together with the variance areas can be computed~ The variance areas exist owing to inaccuracies of the measurements and to the fact that the curve cho$en does not cover·the physical one exactly. In all cases the function is only valid in the measured area.

Note presented to group 0

6

-References

1. Instruction Manual Model 707-120 Crossbar Scanner. Honeywell, Test Instruments, San Diego (California). 2. Instruction Manual Model 630 S Digital Multimeter.

Honeywell, Test Instruments, San Diego (California). 3. Instruction Manual Model 825-C-2 Output Control.

Honeywell, Test Instruments, San Diego (California). 4. J. Mandel

The Statistical Analysis of Experimental Data Wiley and Sons, New York, 1964.

5. Wishart and Metakides

Orthogonal Polynomial Fitting. Biometrika 40 (1953) + 361.

A

ELECTRONIC EQUI PMENT

B

1---1...r READ -OUT

!

__

pu_n_C_h_ed_t_a..:..pr-e---.;N __ Q=-l;...-_...:~s. p ri rna ry . va ria bles

PROGRAMME DIGITAL COMPUTER

•

THEORET ICAl ... - - - 1 MODEL

""--l.;..p_un_c_h.;;.e.;;.d ..,...t

a..:;p...;;.e-=-N;.:Q~2::..-

___

J ..

pr i rna ry +

s~cobnd

a ry

~

varo. I .. PROGRAMME COMPUTOR. - . REGRESSION-PLOTTER

~PROGRAMME

-.. GRAPHS REGRESSION CHOSEN

/ .

+ ACCURACY

CONSTANTS~I

I

8 -PROCESS TO BE INVESTIGATED

I

varia~les

I

I I I digital I informations

..

~ SECOND-SOURCE PANEL 2nd-source command

I

vari~bles

I

I I I I I I I I analogue I I voltages

....

,

INPUT PANEL II II II II II II

JtU

signals + programming cmd.

20 channels~ l~O wires

digital information in parallel form step-scanner cmd r - SCANNER encode cmd. punch command OUTPUT CONTROL information in parallel form information in serie form PUNCHER ;:;: punched ta pe analogue voltage + programming command DIGITAL VOLTMETER