1
Principles of Quality Management and Statistical Specifications
Dennis DvorakFHWA Resource Center
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 1 of 62
2
Topics Why QA? Variability Six building blocks of a Quality Assurance program QC plan Why PWL Purpose of incentives
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 2 of 62
3
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 3 of 62
4
Why QA?
Change in Agency Demographic Agency Personnel Levels Change in Contractor Demographic Rational Assessment of Pay Reduction in Product Variability
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 4 of 62
5
AASHO Road Test
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 5 of 62
6
Construction 1956 to 1958 Highly controlled Specifications based on prevailing state practices Found that variabilitycould not be eliminated
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 6 of 62
7
Sources of Variability
Material Process Sampling Testing
CompositeVariability
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 7 of 62
8
Six Building Blocks Agency Acceptance (QV) Contractor Quality Control (QC/Process Control) Independent Assurance Dispute Resolution Qualified Laboratories Qualified Personnel
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 8 of 62
9
Independent Assurance Assessment of personnel proficiency and equipment Provides independent check on reliability of results of both partners Uses Split sample Not used to make a determination of quality/acceptability of the product
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 9 of 62
10
Sources of Variability
Material Process Sampling Testing
CompositeVariability
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 10 of 62
11
Dispute Resolution Formal system designed to address significant differences between partners data of such magnitude to impact payment Not intended to address day to day issues Required (by FHWA) when contractor results used in acceptance decision Generally uses a “referee” lab
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 11 of 62
12
Qualified Laboratories & Accredited Laboratories
State central labs must be accredited. Private labs conducting dispute or Independent Assurance testing must also be accredited. All other labs must be qualified through a state sponsored program.
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 12 of 62
13
Requirements for Personnel Qualification/Certification
Recommended program guidelines:• Formal training; hands‐on training • On‐the‐job training • Written and performance examinations• Periodic re‐qualification (typically 2‐5 years)• Process to remove personnel performing procedures incorrectly, falsifying statements or data
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 13 of 62
14
Testing Variability Inherent in the procedures and apparatus
Influenced by the technicians
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 14 of 62
15
Testing VariabilityProcedure 95% Lower Limit Test Result 95% Upper Limit
Sieve analysis (% passing ½”) 24% 28% 32%
Slump 2” 2 ½” 3”
Air content 4.9% 5.5% 6.1%
Rodded unit weight for aggregate
114.5 lb/ft3 120 lb/ft3 125.5 lb/ft3
Compressive strength 3,390 lb/in2 3,600 lb/in2 3,810 lb/in2
Flexural strength 602 lb/in2 700 lb/in2 798 lb/in2
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 15 of 62
16
Six Building Blocks (23 CFR 637) Agency Acceptance‐PWL motivates minimizing all variability Contractor Quality Control‐Material and Process Independent Assurance‐Sampling and Testing Dispute Resolution‐Sampling and Testing Qualified Laboratories‐Testing Qualified Personnel‐Sampling and Testing
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 16 of 62
17
Random Sampling(23 CFR 637.205(e)) All samples used in the acceptance decision shall be random samples• All materials will have an equal probability of being sampled
• Removes bias• Reduces potential for fraud
The tests on the first load in the morning are not random!
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 17 of 62
18II - 18
Random Sampling
Simple Random Sampling
LOTLOT
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 18 of 62
19
Stratified Random Sampling
Stratified Random Sampling
Sublot1
Sublot2
Sublot3
Sublot4
Sublot5
LOT
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 19 of 62
20
Non‐Random Sampling Used at start‐up or when process adjustments are made• Example: adjusted air entrainment admixture dosage
• Sample immediately and repeat testing to monitor the adjustment
• Return to random sampling when desired results are consistent
• What should the agency do?
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 20 of 62
21
Boston Central Artery/Tunnel Project
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 21 of 62
22
Quality Control Defined
Quality Control ‐ The system used by a Contractor to monitor, assess, and adjust their production and placement processes to ensure that the final product will meet the specified level of quality.
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 22 of 62
23
Quality Control Uses real time feedback A good QC system:
• Doesn’t try to accommodate Agency requirements
• Implements QC procedures as standard practice
• Isn’t just paperwork…it’s a mindset
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 23 of 62
24
Quality Control
Integrated throughout the organization Not just responsibility of QC personnel Quality can only be achieved by skilled and properly trained frontline workers
Comprehensive, systematic, and continuous approach Proactive, not reactive Focus on preventing problems not finding them Zero Defects philosophy
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 24 of 62
25
Scope of Quality Control Activities Contractor’s QC system should address:
• Materials production • Materials transportation and handling• Field placement procedures• Calibration and maintenance of equipment• Sampling, testing, and inspection to maintain each process “in control”
• Use of QC information to make timely adjustments
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 25 of 62
26
Scope of Quality Control Activities
Inspection is as important to Contractor QC as sampling and testing Necessary to minimize visually detectable problems
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 26 of 62
27
Project Quality Control Organization
No “one‐size fits all” QC organization structure Will vary depending on the size and complexity of the project
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 27 of 62
28
QC Plan A project specific document prepared by the contractor which identifies QC procedures that will be used to maintain production and placement processes in control and meet the agency specification requirements.
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 28 of 62
29
Poor Example Contractor’s QC plan states:
• “Materials, equipment and methods used to construct concrete pavement will conform to Section 501 of the specifications”
• “Quality control testing will be performed in accordance with Section 501.14 of the specifications”
Module 6 ‐ 6
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 29 of 62
30
Improved Example Contractor’s QC plan states:
• Smoothnesso 20’ straightedge advanced at 10’ increments
o Paving foreman will maintain a log of all “events” indexed by station
o Smoothness test results will be provided to the paving foreman for analysis and approval within one hour of testing
Module 6 ‐ 7
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 30 of 62
31
Setting Expectations
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 31 of 62
32
Quality Control Plan Materials production processes Materials transportation and handling Field placement procedures Calibration and maintenance of equipment Activities (sampling, testing and inspection) to maintain each process in control Means to make timely adjustments and corrections
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 32 of 62
33
QC Plan Format and Contents Agency specifications normally identify minimum QC Plan requirements Project specific and reflect actual QC processes that will be applied Written in straightforward and plain speaking manner Comprehensive in the amount of information provided
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 33 of 62
34
Visual Inspection for Quality Control
Visual inspection is also essential to achieving quality Observations and measurement of
• Equipment• Environmental conditions• Materials• Workmanship
Prevention based, not detection based
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 34 of 62
35
QC Inspection of Equipment Ensure that all equipment is in good working condition and is properly maintained Periodic calibration of equipment should also be performed Document prior to production
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 35 of 62
36
QC Inspection of Environmental Conditions
Monitoring and measurements to ensure that the proper physical environment is being provided for:• Storage• Production• Placement• Curing of individual materials or products
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 36 of 62
37
Inspection by Production Personnel Contractor’s should strive to create a “Quality Culture” throughout the company Personnel are aware of how their actions can affect the quality of the product Contractors must invest in the continual education and training of their work force in order to maximize their involvement in Quality Control
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 37 of 62
38
Inspection by Production Personnel
Skilled Laborers and Foremen should be capable of identifying unacceptable materials prior to incorporation in the Work• Should take initial responsibility for accepting or rejecting it
• Should always communicate their findings and proposed actions with their QC Inspectors
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 38 of 62
39
July 2006 motorist fatality Partial collapse of tunnel ceiling on I‐90
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 39 of 62
40
Agency Monitoring QC Activities
Periodic visual observation of QCinspection, sampling, and testing Review of QC records/documents Provide feedback to contractor’spersonnel
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 40 of 62
41
Maine Example Data analysis of materials testing from 1978‐1998 Per cent “passing” tests ranged from 87%‐92% every year Confirms that our industry, without incentives, operates around 90 PWL
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 41 of 62
42
Normal Distribution Previous example provides a picture of variability Pattern formed is referred to as Normal Distribution Properly obtained statistical sample for an entire lot of most construction material will form aNormal Distribution Curve
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 42 of 62
43
This Empirical Rule states that:
–1Sn 1Sn 2Sn–2Sn 3Sn–3Sn
6895.599.7
x
68 – 95 – 99.7 Rule
68% of all possible samples are clustered about the mean within ± 1 standard deviation
95.5% are within ± 2 standard deviations
99.7% are within ± 3 standard deviations
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 43 of 62
44
Testing Targets and LimitsBased on Normal Distribution
Specifications normally identify targets and/or limits for individual quality characteristics
Specification limits should be based on the principle of normal distribution
6.55.5 7.05.0 8.07.56.0Lower Spec
LimitUpper Spec
LimitTarget
Air Content (%)
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 44 of 62
45
Standard Deviation
Sample standard deviation (s)
s =∑(Xi - X)2
n - 1
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 45 of 62
46
PWL Using the sample data, with its mean and standard deviation, we can determine the quality level of the sample
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 46 of 62
47
PWL Estimates the percentage of material within
specification limits• Assumes normal distribution
1.0or 100%
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 47 of 62
48
PWL Efficiently captures mean and standard
deviation in one quality measure
X 3s2s1s-3s -2s -1s
X - means - standard deviation
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 48 of 62
49
Single Specification PWL
USL
LSL
PWLL
PWLUorX
X
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 49 of 62
50
Double Specification PWL
- 100 =PWLL
LSLUSL
PWLU
X
USL
LSL
PWL
+
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 50 of 62
51
Estimating PWL in Four Steps1. Obtain random samples2. Compute
• Mean ( x ) • Standard deviation (s)
3. Compute quality index (Q)4. Convert Q to “estimated” PWL
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 51 of 62
52
3s2ss-3s -2s -s
Compute Q
QL =X – LSL
s
USL - X
sQU =
LSL
USL
X
QL QU
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 52 of 62
53
QL → PWLL
Convert Q to PWL Lookup estimated PWL from Q for a specified
number of samples (n)
X3s2s1s-3s -2s -1s
PWLL
LSL
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 53 of 62
54
Question
Why would you pay more than 100% for a pavement product?
3% Incentive Cost less than 1 year (3% of 30 year pavement design life) Gain: How much longer will it last?
• Better Mixture => Less cracking / reduce early deterioration risk• Smoother ride => less dynamic loading
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 54 of 62
55
What Are You Paying For? Specifications are set at a level for average work
• Average contractor• Increases competition
Low bid system• Won’t get better than that• Do the work the cheapest at the minimum level
Not paying for higher quality
Ford Focus
You get what you pay for!
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 55 of 62
56
Want Higher Quality? Higher quantity costs money
• Benefit of longer life pavement exceeds the initial cost => economic decision
• Better materialo Water reducer that won’t give problems in hot weathero Cost of a larger stockpile for a third aggregateo Cost of a third conveyer and possibly a third aggregate bin
• More personnelo Additional QC testingo Additional QC personnel
• Newer equipment
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 56 of 62
57
How? Quality contractor has control of the process
• Less variability• Less chance of rework
Incentives pay for the extra costs of higher quality Incentives give the bidding advantage to the better contractor• They know they can meet the incentive requirements• They can reduce the bid price by some portion of the incentive
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 57 of 62
58
Why? Higher overall quality with a limited number of incentives• Quality contractor has the quality mindset• Quality contractor applies better process control to all work
• A few incentives provide capital to do all things better Reduces inspection time and cost
• No time spent on disincentives• No time spent on rework
Reduce delays due to rework• Open to the public sooner
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 58 of 62
59
Pay Factor Formula A pay factor formula presents a mathematical equation that typically derives a linear schedule of pay from the PWL AASHTO provides a recommended equation:
Pay Factor (PF) = 0.55 + 0.5 (PWL)(PWL is expressed as a decimal value in this equation)
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 59 of 62
60
Pay Factors Currently many states have pay factors for:
• Strength• Thickness
Future???
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 60 of 62
61
Incentives Incentives and disincentives should be related to performance• Another State ‐ 10‐20% reduction in pay for small deviation from specification
Non‐conforming materials• Strength• Thickness• Air content
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 61 of 62
62
WCPA Annual Concrete Pavement Workshop
Dennis Dvorhak Presentation February 13, 2015 62 of 62
Top Related