這件case,典型的process控管不佳(ISO 7.5.2 & ISO 7.5.1),又因為大家一開始的方向錯誤(認定是day one issue),一直著眼在尺寸設計的精進,所以處理時間有所浪費,其最後的結果便是應急的在一天內將所需的更換品作完並空運到加拿大.最慘的是經過1個月了,目前我還看不到vendor所作的FA.
-----handling process-----
Similar,
我將此case的前後故事整理如下,你可以到韓國的路上多瞭解此case的處理過程:(或是妳可以直接看8/21當天描述的RCA 1-3,8/29的(2)瞭解我目前看到的差異)
發生時間:為8-Aug'2013,Joga通知說一款RO內部的ASO經使用者接上管路使用後發生ASO上蓋脫落導致漏水並損害使用者硬木地板的索賠事件(在KF,組裝ASO後,會使用60psi測試30秒的方式100%檢測是否漏氣).
8/13: Vendor問Joga不良率有多少?不良數量有多少?(當時他們的想法是認定此問題是偶發缺失),打算用replacement方式處理
8/13: 我告訴 Vendor,問不良率與不良數量的意義不大,因為可能有風險的ASO尚未被使用,或是因為偶發異常並未有其他不良(當時我也接受 Vendor提出Joga描述的不 良品是偶發,因為以前真的也沒有發生類似問題,更何況KF使用100%氣測,應該不良數量不會很多,但是,我還是告訴KF如果還有其他同樣缺點模式發生, 那這就不是偶發問題,而是Process某管控環節已經失效或是現階段的Gate無法預防此缺失)
8/13: 我告訴 Vendor,問不良率與不良數量的意義不大,因為可能有風險的ASO尚未被使用,或是因為偶發異常並未有其他不良(當時我也接受 Vendor提出Joga描述的不 良品是偶發,因為以前真的也沒有發生類似問題,更何況KF使用100%氣測,應該不良數量不會很多,但是,我還是告訴KF如果還有其他同樣缺點模式發生, 那這就不是偶發問題,而是Process某管控環節已經失效或是現階段的Gate無法預防此缺失)
8/15:因應Joga要求, Vendor拉力測試10 ASO,標準是Min 50kgf,所有10 ASO全部通過,但Joga認為Min 50kgf的標準過低,要求提高,同時,我也提出此拉力測試只是將快速接頭拉出,並不是真正測試ASO上蓋融接面的強度,並且,測試拉力的意義也不大,因為此缺失如果是偶發缺失,有風險的ASO可能因為數量不多,很難會被檢測出.真正的方向應該要從缺失品的缺點模式反推Process可能導致的根因有那些,然後逐一對症下藥..
8/15:我通知 Vendor再測試30 ASO,並依Joga的要求將拉力設定為72 kgf,並需量測ASO ID + Cap OD以備差異分析
(24-Sep)
8/21:我告訴Joga,應該要求 Vendor review welding process:
8/15:我通知 Vendor再測試30 ASO,並依Joga的要求將拉力設定為72 kgf,並需量測ASO ID + Cap OD以備差異分析
8/21:看過第一個退回的不良ASO(單邊融接,約3/4的面積無有效融接痕跡),我簡單作FTA分析如下:
Status: We had broken ASO that was found unevenly welding and insufficient welding area on the step of cap to result poor structure.
RCA 1: Before welding, the cap onto ASO port was tilting.
RCA 1-1: The tilting could be from interference of dimension deviation.
For Cap
>A: the height of Cap step is too high
>A: the height of Cap step is too high
> B: the OD of the Cap step is too small
For ASO
> C: the depth of ASO port is shallower
> D: the ASO ID is too small
RCA 1-2: The tilting could be from interference of dimension deviation with OVAL issue.
RCA 1-3: The tilting could be from improper compression by welding machines(這是我目前判斷的主因,因為 Vendor OP告訴我,如果融接面足夠的話,通常會伴隨著毛邊需要整修,但8/21我並未看到需要整修的毛邊,反而在9/24(今天)檢查要空運的ASO發現 Vendor OP都在修毛邊,而且融接面(ring extrusion)很漂亮.
(24-Sep) RCA 1-4: The tilting could be from suspected units during set-up or trouble shooting or shut down or breaking time..etc.
Handling procedure to evaluate/find out RCA
1. To review the dimension measurement to see any difference or deviation that could match RCA 1-1 or RCA 1-2
1. Try to replicate the uneven welding by samples from 1.
2. To 100% inspect some welded ASO to evaluate the failure rate of uneven welding.
3. To analyze the process from compressing to welding by slow motion by video to look for any possibility of getting tilting for RCA 1-3 & RCA 1-4.
4. Others..
Remark: This issue may be a complicated case that requires all of us to review from molding to welding with dimension and process study. At least, we may identify what is the risk or possible RCA during analysis process before commenting with any CAPA, this might take some time but it is worthy of that to resolve this kind of issue W/O any more useful data at hand.8/21:我告訴Joga,應該要求 Vendor review welding process:
My recommendation is that even if we ask Vendor to increase pull force test criteria, it is still risky of having same ASO leaking issue as this couldn't be detected by any sampling plan or strict gate.We need Vendor to overall review the process from molding and welding to identify the potential root causes, then to take actions corresponding to the possible root causes. Also, they need to find the different way of inspecting this kind of poor welding as neither current air leaking nor sampling plan could be lucky to find this unit.
8/28: Joga通知有另外2 ASO 有同樣漏水的問題!
8/28: Analysis with Vendor measurement of ASO ID vs Cap OD
1. Measurement, refer to Spreadsheet "distribution" in the attachment:
1.1 To analyze the ASO port ID and Cap OD, non of them meets Vendor DWG Spec. Measurement of ASO Port ID is UNDERSIZED and Cap OD is OVERSIZED that will potential result Cap not sitting properly onto the ASO(being lifted or tilting).
1.2 We could see:
(3) the OUT port ID of ASO#1 is smaller in the groups, and
(4) the IN port ID of ASO#4 is bigger.
Q1: I could not identify CAV# of Cap OD from the attachment, only sample No. We need to know if defective ASO has fixed CAV#.
Q1: I could not identify CAV# of Cap OD from the attachment, only sample No. We need to know if defective ASO has fixed CAV#.
2. Evaluation with welding process on 8/29.
2.1 To sort out at least 100 welded ASOs to check if any uneven welding could be found and what is the defect rate. (Due:8/29)
2.2 I pick up Smaller vs Bigger in both ASO port and Cap, you could refer to spreadsheet "8-28(2)" with Yellow and Blue column. I need them to do fitting test to find if any tilting could be replicated. (Due:8/29)
Q2: Vendor does not provide how many CAV# in Cap tooling, I am not sure if 20 samples is sufficient.
Q3: Another key factor in dimension to result tilting/lifting/uneven issue is Cap step height and ASO port depth that I mark them with red circle in the spreadsheet "ASO DWG" & "Cap DWG". This might also request Vendor QC to do the measuring and monitor in the future.
8/29:撿驗 sonic welding process後,我發現:
(1) 該welder使用的12組夾具,有6組在融接前會有傾斜的現象
(2) ASO融接後,以合模線當Y軸,呈現12點鐘與6點鐘方向有融接(extrusion)痕跡,3點鐘與9點鐘方向約有50%是沒有融接痕跡的,退回的ASO,也是如此,但多了一項tilting現象(如附圖)
9/1:看過vendor的分析報告(他們一直埋首於ASO ID+Cap OD的day one issue),我也建議他們要review welding process.
For welded area, this is a LTB issue that requires KF to review both static and dynamic variations to find KSF.My suggestion from above for KF is not just study only dimensional deviation, the welding process shall be reviewed as well in the purpose of having most effective welded area.If I were KF man, I will draw FTA to find the problem structure and then to check potential risk before commenting something because this issue to me now is still a probability of day one(Cp) issue since several years ago, not easy to take actions for suspected variations because everything could be suspected.
9/18:Joga通知相關人員要將KF RO hold住,並要求KF盡快送 replacement+ reworking instruction in English到WG.
9/23: Vendor returned molding for ASO Cap,Body and started to sonic weld the parts on 9/23.
9/24: I checked the welded and assembled ASO to tell the ring extrusion difference.
In a summary, by my judgment today, the broken ASO could be major from process deviation(insufficient compression or energy on the parts during sonic welding), and the dimensional deviation in ASO ID, Cap OD would be the 2nd RC. But, if vendor could do the welding job well, there should be low or even no risk of seeing broken ASO.
So, please have a review with Vendor ASO when you visit MF, and take the pictures for a discussion if you judge it risky..
So, please have a review with Vendor ASO when you visit MF, and take the pictures for a discussion if you judge it risky..
David
---26-Sep,看過供應商提出的報告後,我請加拿大同事reject,原因如下---
Joga,
I just reviewed the root cause and solution by XX in Vintara for this issue, please reject it if you agree with my below analysis.
By the returned ASO, the defect code was " Unevenly welded"(refer to attached picture) that should require XX to look into the process variation, or interference between Cap/retainer and ASO body before sonic welding. But, XX just review the cap/retainer deviation and to modify the CAV#1,#5 that was not corresponding to the defect(unevenly welding)".
Basically,
(1) XX should explain with any possibilities of causing "Uneven welding",
(2) then define the possible root causes for review.
(3) And, then review the containment action corresponding to the possible root causes.
(1) XX should explain with any possibilities of causing "Uneven welding",
(2) then define the possible root causes for review.
(3) And, then review the containment action corresponding to the possible root causes.
This was a complicated case as most of the possible root causes were not being identified but XX just take modification and strict QC gate(higher pull force criteria) as the action.
The funny thing is that we should return ALL ASO if they judged CAV#1 & #5 of retainer is risky because pull force is lower than average. But, on the opposite, even the pull force with CAV#1,#5 of retainer could meet 72 kgf. And, NO one could make sure returned ASO was welded with risky retainer(CAV#1 or CAV#5).
In a summary, the root cause and solution from KF does not make sense to me at all.
Thanks,
David
---27-Sep,加拿大同事對我看法的回覆,要求供應商修正報告內容---
Rework Instructions: Please address follwoing concerns::
1. Root cause Analysis has not done properly on the returned units whether it is systematic or procedural errors such as operator negligence etc. (P3_Failure mode analysis & P6_ Root cause analysis)
1. Root cause Analysis has not done properly on the returned units whether it is systematic or procedural errors such as operator negligence etc. (P3_Failure mode analysis & P6_ Root cause analysis)
2. vendor has not mentioned about containment actions including affected part #'s , WIP, their Warehosue stock, on ocean etc., even though WG has quarantined all suspected models.(P4_Containment action)
3. There is no mention to control the implemented actions for review in future. All the test data is to be recorded In Process documents. (P7_Decision making(Preventive action)
--02-Oct,加拿大同事告訴我又有一個不良品----
與OEM主管討論後才明白原本零件A的tool core於今年(2013)曾經更換過,因為並未將融接面的尺寸納入重點管控,所以導致融接不足的可能性的確是存在,並且也經過DOE驗證.
現在的問題是,已經出貨並在消費者端使用的產品該如何解決,我明白加拿大同事打算recall,但OEM主管解釋不良的比率不會太高,且當不良品發生時間會因為使用時間長短而有所影響，推估已使用產品的風險應該較低，再搭配品質成本計算,判斷不需recall是比較划算的辦法。
--03-Oct, Process verification---
Hi, All
Here is the verification for your information:
1) For defect code: Insufficient welding
1.1) Verification on 3-Oct at Welding process:
(1) The welding extrusion is good and acceptable
(2) I did not see the control range for welding parameter(Needs to be revised and update the action in Vintara)
1.2) Pull/SI test verification
(3) All the test with corrected ASO,Retainer is acceptable. But, it is better if vendor could tell why the pull test was so good on 26-Sep.
(4) By the test samples, I could see the welding condition is much better than before, refer to <<100313_KF Sonic Welding_pull test samples>>.
2) For defect mode: Unevenly welding
2.1) Verification on 3-Oct at Welding process:
2.1) Verification on 3-Oct at Welding process:
(5) The shifting at some welding fixtures is still being found during welding, refer to the video, please fine adjust the concentricity.
(6) The tilting retainer/cap is still being found, refer to the video, this may be a task for vendor to study but I am OK so far because vendor had improved to have more welding amount by modifying both ASO and Retainer/Cap dimension.
(7) Remark:
R1: We have one unevenly welding ASO so far, this failure could be combination of two factors: A_tilting cap & B_Insufficient welding amount.
R1: We have one unevenly welding ASO so far, this failure could be combination of two factors: A_tilting cap & B_Insufficient welding amount.
R2: It is recommended that KF develops self-inspection by operator to check unevenly welding.
Judgment: By the corrected ASO & Cap and welding process, it is acceptable.
If Joga has other question or concern, I will check with you accordingly.
全站熱搜
留言列表
Well, what is the proposal? LEAN Production (2)
