How the Alignment System Works – Choosing Reference Points/Datum
The most important step in aligning any machinery is determining what to align it to! In other words, what are your reference points or datum? For a centerline, such as in bore alignment, or a linear axis, such as in machine tool alignment, you will need to find 2 reference points to use for aligning or “bucking-in” the laser. Once the laser is “bucked-in” (aligned) to the reference points, you can then start measuring the other points in the linear axis (or bores) to see if they are aligned relative to the reference points. For a surface, you will need to find 3 points. Alternatively for measuring flatness, a common reference is using earth level. Our L-730/L-740 Series lasers have level vials built into them, so either method can be used to measure flatness. Bad Reference Points, Bad Data A basic rule to remember is if you have “bad” reference points, you will have “bad” data, so it is really important to think through what you want to use as a reference. You don’t want to waste a lot of machine downtime trying to realign a machine to bad reference points when the machine may already be in alignment! Level as a Reference Not Always Reliable Another point to consider is earth level is not always a reliable reference to use for measuring flatness since you could have a machine bed that is not level but is flat. In this case, you will waste a lot of time releveling the bed when you don’t have to. For example, say you have a flatness tolerance of .005 in. (0.13 mm) TIR for a surface and you used the levels on the laser to measure the bed. Then let’s suppose at the farthest point from the laser, you found a flatness error of .020 in. (0.5 mm). This would most certainly appear to out of tolerance, or does it? It could be the case that the surface is not level but it is flat (see the Why it’s Best to Use Data Analysis graphic below)! Therefore, to ensure that you are not using a “bad reference”, we find it’s best to use 2 reference points instead of the levels. So let’s look at the same example but instead we’ll use reference points on the surface itself and buck-in the laser to the 2 end points. Then we measure the points in between to look for flatness errors. In this case, it could easily be that you find a max flatness error of only .003 in. (0.076 mm) relative to the reference points. Well this would be well within your tolerance and you would save yourself a lot of time trying to relevel the bed when it was already flat! This happen way more than you think, so always use reference points if you can!
Reference Points/Datum
Points chosen on a surface, in the travel of a linear axis or in a bore that represent the starting point (reference) for which all other points on the surface will be compared. Reference points form a line or a plane that is called a datum. For bore, spindle and linear travel applications 2 reference points are needed to establish a datum - for surfaces or reference planes, 3 reference points are needed.
SystemRecommendations All Hamar Laser straight-line and scanning laser systems
Choosing Reference Points in Laser Alignment
Application Note
On Pr Convebenchmreferenreferenalignmbecausfloor, nickedconcreitself. benchm Use a Fortunabilitynecess On M Machifor alirequirebeforeScan®the hoplanes Linea In genreferenflatnesmess uis conttravel.but it ifinal ranalys Alway It’s alwbecaususuallypart tolong. For exaxis anreferen
rocess Mills:
entional methomarks (monumnces. Howevence roll provid
ment and resulse benchmarkare rarely cov
d. More imporete and rarely Unless checkmark will prob
reference roll
nately, our L-7y to use a refersary.
Machine Tools
ine tools are ognment. Typie multiple refe
e starting any a® laser, for exaorizontal (red) s (either green
r Axes vs. Ma
neral, it’s bettence versus thess errors alongup your referetrolling the cu. Of course, yis best practiceeference and t
sis of the data
ys Use Longes
ways a good idse it gives youy a bad idea to
o measure a su
xample, let’s lond a 30-foot (nce to measur
ods of roll aligments) at the ser, we stronglydes a more accts in better alis are usually s
vered, and are rtantly, they mhold their pos
ked every timebably result in
and you will
742 and L-732rence roll or to
:
ne of the mostically they haverences that nealignment jobample requiresplane, and 2 for yellow).
achine Tables
er to use the me table. Tablesg with many dnce. Also sin
utting head, it’you can use thee to use the mthen us the lea(see below) to
st Distance be
dea to use the u the best anguo use 2 referenurface or linear
ook at machin10 m) X axis. re squareness o
gnment usuallyside of the macy believe that curate referencgnments. Thi
set in a thin coroutinely run
move with theirsition relative te they are usedn significant al
never have thi
2 Roll Alignmo use benchma
t challenging ve tight toleraneed to be cons. Using our Ls 5 reference pfor one of the v
s
machine linear s can have sig
dings and dentsnce it’s the lines best to focuse table as a sta
machine’s lineaast-squares, beo remove laser
etween Refer
longest axis aular accuracy. nce points thatr axis that is 3
ne tool with a 5 If you used th
of the X axis,
y use floor chine as using a ce for roll is is oncrete over and r slab of to the mill d, lignment error
is problem.
ment systems ofarks if absolut
applications nces and sidered
L-743 Triple points, 3 for vertical
travel as a nificant s that can ear travel that s on the arting point ar travel as est fit r slope error.
rence Points
as the referenc In other word
t are 5 feet (1.0 feet (10 m)
5-foot (1.5 m)he Y axis as ththen you wou
rs.
ffer the ely
ce ds, it 5 m)
) Y he ld
Horizontal Floorr Mill with 30 ft X TrTravel and 5 ft Z Travvel
need t So let’your X(.001/5 But if end of Least- In metanalysproceesoftwastraigh This stthat ondrawinsetup p It workstraighplane line (oanalyzthe rawslope evalue negatiTotal Itrue fl Best-F By usislope ecase othe Y analysfrom tbe detData A
o buck-in (alig
’s say you bucX axis, you wo5*30 = .006 o
you use the Xf the Y axis wi
-Squares, Bes
trology, it is resis be performeding to the alare uses a leashtness and flat
tep can save cnly has a coupng a wrong coproblem.
ks like this: ahtness (or flatnfor a surface)
or plane) will hze flatness/straw data and theerror of the be(relative to theve value (absoIndicated Runlatness error o
Fit Analysis R
ing the Best-Ferror caused b
on the machineaxis as the ref
sis, then the lathe data and thermined (see tAnalysis).
gn) the laser to
cked into the Yould have an laor 0.025mm/1.
X axis for yourill be .00017 i
st-Fit Analysi
ecommended ed on a machilignment staget-squares, bestness data.
countless hourple of bad poinonclusion from
a Best Fit line ness) data for that “best fits
have a slope toaightness datae points are coest-fit line. The best-fit line)olute value) tonout (TIR) flatf the surface.
Removes Setu
Fit analysis, weby a laser setupe described abference, if we ser slope error
he true flatnessthe graphs in
o the Y axis at
Y axis to .001 aser-setup erro52m*10=0.16
r reference andn.( .001/30 * 5
is
that some sortine's axis befoe. All of Hamat fit (Best Fit)
rs aligning a mnts, or perhapsm the data due
is calculated fa linear axis (” the data. Tho it that can be. This slope is
orrected to remhen the maxim) is added to mo produce the ttness error, wh
up Errors
e can also remp problem. Soove, where weused the best-r would be rems of a surface wWhy It’s Best
t a 5-times hig
in (0.025 mmor of .006 in. d
6 mm). We ca
d align to the s5 = .00017 or
t of data ore ar Laser’s analysis of
machine tool s from to a laser
for a set of or a best-fit
he best fit e used to s applied to
move the mum positive maximum total error or hich is the
move the o in the e used -fit moved would to Use
gher accuracy
m), and then yodue to the laseall this the “tai
same .001 in, t0.025mm/10m
Hamar’s M
than if you ha
ou measured ther having a sloil wagging the
then the laser-m*1.5 = 0.004
Machine Tool Geo Somachine
ad used the X
he X axis. Weope of .001 in e dog”! Not go
-setup (slope) 4 mm). Much
oftware automaticallye tool straightness da
axis.
ell at the end o5 feet ood.
error at the better!
ly uses Best-Fit analydata
of
lysis of
Squar Since using the topIn othedetermpick 2normamake the Beyou a Best Fbest anSoftwa
reness/Paralle
most axes of many 2 points op and bottom per words, pick
mine squarenes2 middle pointsally there are sit very hard to
est Fit analysisslope of its Be
Fit line for the nalysis of the are does this).
elism Measur
machine toolson an axis as apoints on the Zk the top and bss to X and it s and it may bstraightness ero pin down thes of the straighest-Fit line forX axis straighsquareness be.
rements Requ
have straightna measurementZ axis), could bottom points could look to
be in tolerancerors in the Z ae true squarenehtness data forr Z, and then chtness data, thetween the 2 li
uire Best-Fit A
ness/flatness et for squareneslead to a poorof the Z axis tbe out of toler. This is beca
axis data, whicess error. Butr Z axis, whichcompare this toen you will hainear axes (all
Analysis
errors, ss (such as r analysis. to rance, but
ause ch may t if you use h gives o slope of ave the Hamar
Using a Best-F
Using a B
Fit Analysis for Mea
Best-Fit Analysis for
asuring Squareness
r Measuring Paralle
of Linear Axes
elism of Linear